Park Air Electronics 3030

Last year one of my clients tentatively posed the question, "Would you be up to looking at some Park Air Electronics Air-Band transceivers?". This was followed by, "I'll send you a link, Getting hold of the manuals might not be easy. Let me know what you think."

The link took me to a website of a company which appears to specialise in reselling high-end equipment, at high-end prices. My first impression was that it looked like 90s Plessey kit. I did some research into Park Air Electronics ... Fascinating! The company was started in the late 1960s by Fred Parker (G3FUR) who started out converting Japanese AM/FM portable VHF radios to receive Airband AM, which were often sold at air-shows. For a brief insight into the early days of Park Air, see here.

Fred Parker's enterprise grew to become a well respected manufacturer of high quality Air Band equipment. The series 3000 appeared in the early 1990s (I believe) and I am told that it was considered the Rolls-Royce of Air Band equipment at the time. Although designed as stand-alone transceivers, they were also supplied as part of the BAe Systems AR-327 Tactical Radar System, sometimes known as the T101, ... and my client had his eye on several that were being decommissioned, and were thus up for sale. Each AR-327 actually includes no less than six PAE 3000-series transceivers ... two 3010 VHF transceivers and four 3030 UHF transceivers.

NOTE: It must be pointed out that although owning one of these PAE 3030 transceivers is perfectly legal. However, connecting it to an aerial and pressing the PTT is decidedly illegal. To that end, this article is purely descriptive and observational.

In November last year (2025) I took delivery of two wooden crates, each containing a PAE 3010 transceiver. I was also sent a beautiful set of manuals, Operator's and Technical ... more on them later. I had already been informed that both 3010s were 'unserviceable'. One had an RAF fault report taped to the top cover. The description of the fault was somewhat cryptic, stating that there was a tone on the Guard Frequency (121.5MHz) and the front panel didn't work. All series 3000 transceivers can be fitted with a dedicated Guard Receiver, but neither of them included this feature. If the other 3010 had originally sported a fault report it had since been removed. This one turned out to be fully working. On first powering it up the front panel controls were non-responsive, but this was simply because it was in Remote mode, and a couple of specific button presses changed it to 'Local'. I gave it a full checking over and found it to be fully functional.
The other transceiver did produce a loud tone from the loudspeaker, but on ALL frequencies. It too was in 'Remote' mode, and once changed to 'Local' the noise stopped. It had several other faults which will be covered in a future write up.

There is almost certainly a very good reason why both these transceivers were in 'Remote' mode. I will use Edinburgh Airport as an example. Although Edinburgh Airport has a very attractive 'thistle-like' Control Tower. The ATC towers are not on-site, but several miles to the East, on the top of Corstorphine Hill. Thus whatever transceivers are employed, they too are on the hill and operated remotely. The fact that both 3010s that I had received were in 'remote' mode and both with seemingly unresponsive front panels, does suggest that they had come from a remote site. And it does make me wonder if first-line repair technicians have mistakenly thought that the front panel was faulty. All Series 3010s (and 3030s) store the last used front panel settings in non-volatile memory.

The reason for mentioning the 3010s is to highlight what I might call a Park Air curiosity. See the photograph at the top of the page, and the legend top right which reads 'SERIES 3000', and below that you may be able to make out a hand-written 3030. Now look at the photograph above, of the two 3010s, and you will see that in the same place, someone has written with a black marker, '3010'. Were Park Air saving money by using a generic 3000-series front panel? We will probably never know, as Park Air Electronics no longer support the Series 3000, evidenced by their refusal or inability to provide manuals or any information whatsoever on the type. Around the time that the Series 3000 was in production Park Air was part of the Westinghouse Electronics Systems Group, having been bought in 1990. In 1996 Northrop Grumman purchased Westinghouse and thus became the parent company. Since June of 2023 Park Air Electronics has been part of Indra Group. So it isn't really surprising that information on a 30 year old transceiver like the series 3000 is hard to find. On the other hand however, my client is a man with many contacts, and through one of these he was able to borrow a set of manuals and have them professionally copied.

In January, the two 3010s were crated up and returned and I was assured that the crates would return to me with another two 3010s for repair. However, due to a mix up, what turned up was a pair of 3030s ... The UHF version of the Series 3000. Both were accompanied by RAF fault tickets (MOD Form 731). One of them was in very poor mechanical condition and looked like it had been dropped on its rear ... several times! The fault report for this one was in shocking condition too, looking like it had spent a long time in a very damp place. As with the previous 3010s, the fault tickets were similarly vague ... Not Transmitting or Receiving ... Doesn't turn on + Front Panel doesn't work properly.

I checked the mains fuse on the 'battered' one and found it to be vaporised. The inside of the glass looked like it was coated in thermionic valve gettering. such was the damage to the rear of this transceiver, that I had to spend time straightening out the side panels before I could get a look inside the PSU compartment. Once inside, I noticed that several Earth wires had been deliberately disconnected. Clearly someone had been chasing a fault ... no doubt related to whatever had turned the fuse wire to plasma! I put this transceiver aside and turned my attention to the other one ... the one in the photograph at the top of this page.

I was prepared to accept 'Doesn't turn on' as a valid fault, it being less cryptic than 'Front panel doesn't work properly'. It did turn on, but after a short while it switched off, then back on again. The bulk PSU in the 3030 and 3010 is a fairly straight-forward affair enclosed in the rear compartment ... see below.






The transceiver can run from an AC Mains supply or from an external DC source. The PCB on the right in the photograph above provides the supply voltages for all the internal circuitry with the exception of the RF Power Amplifier module (or modules) ... the +20V or +25V for these is handled by a separate regulator board which is mounted on the left-hand heatsink. (see the photograph on the right).






Space is a premium in this compartment, especially where the toroidal mains transformer is concerned. The damage to the switch was the result of the loops of wire between S1 and S2 pressing against the transformer. When you look at the placement of the two switches in relation to the transformer, it is clear that they are directly opposite the hole in the centre of the toroid. It may be intentional, although not documented, but to avoid compressing the wires between the switches and the transformer, the loops of wiring between the two switches should be perpendicular to the rear panel such that they fit into the hole in the toroid when the panel is folded up and secured.



The side-walls of a typical Series 3000 transceiver are hefty heatsinks. Each transceiver can be considered comprising eight modules with Module-1 being the heatsink forming the left side of the equipment. This supports the PA module regulators, VSWR bridge and output filter. Module-2 is the Transmitter Control Module. Module-3 is the Synthesiser Module. Module-4 (if fitted) is the Guard Receiver. Module-5 is the Remote Control Decoder Module. Module-6 and Module-8 form the actual receiver, with Module-6 comprising the front-end and 1st-Mixer stages, and Module-8 comprising the IF stages and receive audio processing. There is no Module-7, although there is provision. The RF Power Amplifier (RFPA) can be considered Module-9 and is mounted on the heatsink which forms the right-hand side-wall.


Above: This is the BITE (Built In Test Equipment) Assembly PCB. However, it also serves as the interface between the Front Panel and the Synthesiser, IF/AF Module and the Transmitter Control Module. The audio amplifier which drives the small front-panel loudspeaker is on the extreme right of the board. I'm not really sure where to start with this board, or boards. The BITE side of things is quite sophisticated. Unlike RACAL's approach to BITE where voltages and signals are all referenced to a single voltage, Park Air's BITE is based around an MC14433P, an A to D Converter, technically a DVM chip. Everything on the board is under the control of an 8031 8-bit microcontroller (IC17). Whilst the description in the manual erroneously calls it a microprocessor, it is listed in the Items List as an MRO-Controller. I think MRO in this case might be a misinterpretation of a badly written 'micro'. The firmware is stored in an 27256 32K x 8-bit EEPROM (IC15), and a 6116 provides 2K x 8-bits of Static RAM (IC14). Although specifically described in the manual as the BITE Assembly, the two banks of 7-segment displays on the front-panel are also under the control of the Microcontroller (IC17). This level of integration is rather elegant but is let down in a way by the bulky bundles of wires that connect to the headers towards the bottom-right of the board. These can be problematic, and care must be exercised when re-fitting the front-panel. Increasing the space behind the front-panel by 10mm would have been more than adequate.

My HP8920A is set up for verifying the performance of the 3030 on transmit and receive. Whilst the receiver tests returned figures well within specifications, although the transmitter modulation tests were likewise in spec., my Bird power meter was showing a significant reduction of power between 225MHz and 300MHz ... only 5W!! The Spec. is for 40W on AM and 60W on FM. I then noticed something odd. As with the RACAL 3700 series, the Park Air 3000 series provides two types of BITE, Continuous and Interruptive. The latter is invoked when the TEST button is pressed. This must only be done when the Antenna port is connected to a suitable antenna or dummy load. With the frequency set to 399.975MHz, I pressed TEST and a few seconds later the Monitor Display displayed PASS. I got the same result at 225MHz, but between 260MHz and 295MHz it displayed BITE and on pressing MEM, it displayed 8 - 2, indicating an audio level fault on the IF/AF module (Module-8) ... but why only between 260MHz and 295MHz? And if the power o/p was as low as 5W at 225MHz, why was it passing BITE at that frequency? The cause of the audio level fault turned out to be on Module-6, the Receiver RF Module. (see below)


In the PAE 3030, the Receiver RF module consists of a bank of 5 selectable Band-Pass filters (see below), an RF amplifier followed by a mixer which provides an IF output at 70.7MHz. This is followed by a band-pass filter at that frequency. This signal is then fed to the IF/AF module (Module-8).


The five PIN-Diode-selected Band Pass filters on the input to the 3030 receiver RF Module are as follows:
225MHz - 259.975MHz
260MHz - 294.975MHz
295MHz - 329.975MHz
330MHz - 364.975MHz
365MHz - 399.975MHz
After determining that the fault was on the RX module and subsequent exhaustive tests failing to link the BITE fault to the PIN diodes, I eventually removed the board from the chassis and turned it over and found that the output from the BITE Oscillator had not been soldered to the PCB ... a manufacturing error. How did that get through final test in production? I soldered the inner of the coax to the PCB and reassembled the module. BINGO! Interruptive BITE passed at all frequencies ... by that I mean that it now passed between 260MHz and 295MHz.



I was curious as to why BITE passed at 225MHz even though the power output was as low as 5W, but my focus was now on finding the reason why the power output was so low at 225MHz ... that was more important.


Access to the RF Power Module is achieved by removing about twelve screws and then folding the right side-panel heatsink down. Some of the screws are not easy to access, but fortunately the manual does describe where they are and which ones to remove.




Left: This is the 1st Driver Stage in the RF Power Amplifier ... and rather innovative it is too. This stage employs three UHF MOSFETS, two MRF314s followed by an MRF316. The Bias supply for these three devices is also the AM Modulator. This is made possible by the fact that MOSFETS are voltage-controlled. Here is how it works. I'm essentially 'telling it to the duck' here, in the hope that I will understand it myself! The manual not being very helpful. This little board also incorporates an RF drive detector and a temperature sensor. If for some reason, the PA Block heatsink gets too hot or the RF Drive (Synthesiser output) fails, the PTT line will be inhibited and the Power and Modulation Control Line will remain negative. There are other situations external to this board that will also inhibit PTT. Assuming PTT is asserted, the voltage on the Power and Modulation Control Line will go positive. How high it will go is dependent on a look-up table in the EEPROM on the Synthesiser board. If the selected frequency is 399MHz the voltage will be about +7V. This switches on TR5 and TR4 which passes +20V to Zener D3 which by way of RV1 sets the bias voltage on the Gate of TR1, ideally setting the quiescent current in TR1 at 90mA. The Power and Modulation Control voltage is also applied to RV2 and RV3 which set the bias voltages on MOSFETS TR2 and TR3. Since this voltage is by design variable, the bias voltage to these two MOSFETS varies and thus serves two purposes. It allows the output power of the PA to be controlled , and if the selected mode is AM, the modulation signal will be superimposed on this, and thus modulate the output accordingly.




It is vitally important that the amplification achieved in the RFPA is linear. Thus, to aid setting the quiescent current for each of the MOSFETS in this stage, there are three 1-ohm resistors, one in series with each Drain. Two are visible in the photograph above, the third is underneath the large electrolytic capacitor. Thus, millivolts dropped across the resistor equates to the Drain-current in milliamps. Setting a Drain-Current of 90mA for TR1 suggests this stage is running Class AB. Setting the Drain-Current for TR2 and TR3 to 300mA each suggests that both these stages are running Class A.


Left: This is the second driver stage in the RF Power Amplifier (RFPA). It comprises an MRF392 operating in Class-AB, which Motorola describe as a Controlled 'Q' Broadband Push-Pull RF Power Transistor. It is formed from two matched bipolar transistors in a single package with common Emitters and separate Bases and Collectors. This circuit proved to be the cause of the low power at 225MHz. There is no mention in the manual as to what the quiescent current in both halves of the MRF392 should be. The transistor and two diodes adjacent to the RF device serve as a simple bias supply, with approximately 0.8V being applied to the Bases of the MRF392.

However, with no RF drive, the voltage on the Bases measured about -3.2V. Given that there are no negative supplies to the RFPA, something had to be oscillating. I suspected the 2u2F tantalum bead ... untrustworthy things, tants. On replacing it with an aluminium type, the bias voltage on the Bases was as it should be. I tested the transmitter and found the RF output greatly improved, although still below spec at 225MHz. It was now delivering 20W ... which is better than 5W, although 40W would be perfect!

The MRF392 is an extremely robust device offering a gain of 10dB and fully capable of delivering 125W. There is no mention in the manual of what power this stage delivers, but I would estimate it to be somewhere between 5W and 10W. Thus I suspected the MRF340 pass transistor in the bias circuit. Observing it closely, there was a distinct 'whiff' of heat from it, and on closer examination it appeared to be cracked. My temperature probe confirmed the presence of heat by indicating +65C! Ouch! What did strike me as odd was the lack of a heatsink. The device appeared to be flat onto the Duroid.



I de-soldered the MJE340 from the board and it fell apart! (see below). The middle photograph below is also very revealing. The Duroid material under the transistor has been damaged by the heat. No attempt has been made to wick any heat away from the pass transistor. The TX/RX duty-cycle in a transceiver is something that is often taken into account when considering thermal build-up. But in this case, since the DC supply is permanently connected to the RFPA, the bias supply on this board is permanently delivering current. The other thing to consider here is that on AM, the supply to the PA is 25V. When FM is selected, it is 20V.

There is nothing in the transceiver manual regards the quiescent current in this stage, so I inserted a 0.33R resistor in series with the Collector supply. When AM is selected (no RF drive), the drop across the resistor was 46mV which equates to a quiescent current of 140mA (70mA per half), which sounded right for its role as a driver stage. Given that the voltage on the Bases is in the region of 0.8V, the drop across the MJE340 is therefore 24.2V. If the hFE of the device is 50 (reasonable estimated average), then the combined Base currents could be around 2.8mA. BUT, each Base supply is shunted by a 10R resistor. These are effectively in parallel, so we need to take the current in each of these into account as well. In this case, that equates to an additional 160mA. So the total current drawn through the MJE340 is in the region of 163mA ... and since the voltage drop across the device is 24.2V, the power dissipated in the MJE340 is in the region of 4W ... and that's hot enough to burn your finger!

This is clearly a serious design error, or at least a design over-sight. It is likely that the voltage delivered to the Bases (0.8V) was used when calculating the power dissipated in the MJE340 rather than the voltage drop across it. At this point it hit me that ALL PAE 3030s are likely to have this fault ... hmmm?

There is precious little room in this compartment to fit a heatsink. You could cut a hole in the board to provide contact with the metalwork but that would be messy unless you removed the RFPA from the heatsink ... and given that the entire RFPA would have to be removed ... and then maybe even separate the 2nd driver module ... nope, too complicated! ... forget it. I decided to simply move the MJE340 onto the side wall, as in the photograph below-right. And it works. The TO-126 package does not require a plastic insulator through which to fit a mounting screw, and a standard TO-220 mica thermal washer/insulator can be used to isolate the Collector tab from the side-wall. The temperature of the device does not rise above +30C and the RF output was now within spec. The legs of the MJE340 are even long enough so as to not require additional wiring.



The output stage of the RFPA comprises two identical MRF392 push-pull amplifiers, essentially identical to the 2nd driver stage but with no active bias circuit. DC-wise, all four Bases are each shunted to ground by a 10-ohm resistor, so I believe this final output stage is operating in Class-B.

Initially, the Technical Manual appears to contain everything one would expect of such a publication. However, as is not uncommon with manuals for complex equipment, it would appear to have been written by engineers who designed the Series 3000 for engineers on the project. This is more an observation than a criticism. However ... There is a thing that is frequently referred to as a mod-strike label. It is simply a matrix of numbers, where each number denotes a level, or modification status. Each sanctioned modification is normally referenced in an addendum in the technical documentation. Thus, anyone who is called upon to repair such equipment will automatically look for the mod-strike label and note the numbers which have been struck-through. This will enable them to reconcile variations from the norm with the appropriate addendum. That's how it is supposed to work.

Having repaired several faults in this particular PAE 3030, I was not happy with the disparity between what my Bird Power Meter was displaying and that which was displayed numerically in the Monitor Display. I figured that such a thing would be covered in the manual, in the chapter describing the Transmitter Controller, or in the Alignment Chapter further on in the manual. Not so, or at least it isn't obvious, if it is. The manual does describe a loop whereby the level of output from the RF Power Amplifier is maintained/controlled. This involves the VSWR bridge and elements of the Transmitter Controller module. However, the manual stops short of specifically describing how the key elements in the loop interact. What is obvious is that IC31 has something to do with it all ... and that's where the manual falls over.

There is a line in the manual, in Section 4, Page 14, para 112 which reads 'locate the test points TP10/TP44/TP45, jumpers J4/J6. and potentiometers RV11 to RV14. Set J4 to the off position and set J6 to the on position'. I found all the test points and the two links, however, on the board that I was looking at, there was no RV11, RV13 or RV14. I only had RV12 ... and not only that ... whilst the manual showed IC31 as 14-pin MC1495, I was looking at an 8-pin AD633. Both chips are Analog Four-Quadrant Multipliers. Given the simplicity of the latter, I can see why the MC1495 was replaced and the board re-designed. The setup procedure in the manual is based on the Issue-D version of the board, whilst the one that I was looking at was an Issue-G board.


My client asked his source if they had a later version of the manual, one that might refer to the later version of the Transmitter Controller. Sadly, he did not. Towards the end of 2025, on noticing that some of the transceivers had their mod-labels struck off up to 5 or 6, 'we' had asked Park Air Electronics if they might be able to provide us with a list of changes relating to the numbers struck off on the label. That request had also drawn a blank. So, there was only one solution ... reverse-engineer the Issue-G board ... or at least the circuitry around the AD633.


Notice that the 'logic' relating to J4 appears to be reversed on the Issue-G board. Incidentally, jumpers J4 and J6 appear to be for circumventing some internal monitoring interlocks. For instance, if the RF input to the Power Amp module is disconnected, the PTT circuitry is automatically disabled. Similarly if the 6-pin (3-wire) connector to the Power Amplifier is similarly disconnected, the PTT will again be disabled. What I found when following the procedure in the manual was that regardless the settings of J4 and/or J6, if the drive to the PA was disconnected and/or the 6-pin connector was disconnected, the system automatically entered a BITE notification state which rendered the procedure impossible. I tried both version D and G controller modules and got the same result for both. I suspect transceivers with the Issue G board have different firmware. I should probably investigate this.

I was still unhappy with a disparity between the Forward Power displayed on the front panel and that displayed on my Bird Power Meter. 'Going by the book' was getting me nowhere. So I decided to go back to basics. As far as I could see, there was no way of breaking the power control loop without disabling the transmitter. I followed the manual as far as ensuring the correct voltage on TP10 on the Transmitter Control module. Having ascertained that at 399MHz, the voltage on the Modulation control line was +7V, I then set the frequency to 399MHz but disconnected the Synthesiser Drive to the RFPA. I disconnected the Modulation Control line from the RFPA and applied +7V from a DC power supply, then switched on the Transceiver. With no RF into the RFPA, PTT would be disabled. With the Mode set to AM, I verified that +25V was being applied to the RFPA. I was now in a position to set the Quiescent Currents for the three MOSFETs in the Pre-Driver stage. All three were significantly out of spec. TR1 was then set to 90mA and TR2 and TR3 were subsequently set to 300mA.

I then switched off the transceiver, reconnected the Synthesiser Drive and the Modulation Control line. I then switched the transceiver back on and set it for 300MHz AM. I also set the Forward monitor potentiometer on the VSWR bridge deliberately fully anti-clockwise and set the front panel SET POWER control fully clockwise. When I pressed the PTT, the power on the Bird Power Meter was way down low, which I expected. I then turned the forward power potentiometer on the VSWR bridge clockwise until the Bird Power Meter read 40W. The Forward Power reading on the front panel was now reassuringly registering 40W. On FM the Bird measured 60W and the front panel did likewise. Only below 250MHz did the parity between the front panel display and the Bird Power Meter vary by about 10%.

An interesting feature of the Transmitter Control Module is the inclusion of an RF Clipper. The prime purpose of this is to prevent any overmodulation of the transmitter. It also ensures a more constant level of modulation and thus readability. This is achieved by employing compression on the audio from the microphone. However, a feature of RF-Clipping is that it also guarantees that any harmonics produced by the compression process are well outside the audio pass-band of the transmitter. In this case the audio input (300Hz to 3400Hz) is transformed into an SSB signal at 60KHz which is then clipped/compressed. This is then demodulated back to base-band (300Hz - 3400Hz).

Module-3: The Synthesiser Module, pictured below, is typical of such designs first seen in the 1970s through to the late 1990s when DDS (Direct Digital Synthesis) became more popular. It provides the drive signal for the RFPA and the signals for the receiver 1st and 2nd Local Oscillators. The synthesiser module in the PAE 3000 series is common to all models and has its own dedicated processor board based on the 8031 Microcontroller. The firmware for this module is on 27128A 16K x 8-bits EEPROM whilst a 6116 provides 2K x 8-bits of Static RAM. The actual synthesiser itself comprises two boards within a screened compartment. In the photograph below, the centre section contains seven selectable VCOs covering 100MHz to 471MHz. Control and selection of these VCO is via the Synthesiser board on the right. The description in the manual makes the whole frequency generation process sound more complex than it actually is. Even the block diagram will give you a headache. The schematics for all three boards are on three separate sheets. The VCO board schematic is self-explanatory and the schematic for the Synthesiser is easy to follow since it is laid out functionally.

The heart of the synthesiser is based around two 'dedicated' chips, an HEF4750 Frequency Synthesiser and the accompanying HEF4751 Universal Divider. The output of the HEF4750 drives a Phase Locked loop which in turn provides the tuning voltage for the VCOs. A sample of the latter is fed back to the HEF4751 via several ECL dividers, completing the loop. When Frequency Modulation is selected, the deviation is applied as a signal superimposed on the VCO tuning voltage.


Module-4 is the Guard Receiver and if fitted, incurs an expected drop in receiver sensitivity of 6dB. When not fitted, the appropriate connectors on the back-plane edge connector are linked together. I would envisage that most, if not all commercially operated transceivers would be fitted with the appropriate guard receivers. This transceiver was part of a BAe Tactical radar system so was not fitted with a guard receiver.

Module-5 is the Remote Controller Interface; designed to allow remote control over 600-ohm lines. As I don't have the PAE 3000EB Remote Control Unit, I have no way of testing it. What I can say is that as with the BITE Assembly and the Synthesiser, the Remote Controller Interface is host to yet another 8031 Microcontroller! The firmware is on a 27C512 64K x 8-bit EEPROM while a 6116 provides the usual 2K x 8-bits of Static RAM. There are also a couple of 22V10 PALs (Programmable Array Logic).


There is no Module-7.

Module-8, the IF/AF Module is the second half of the receiver and includes the 2nd mixer which takes the 70.7MHz IF down to 10.7MHz. This module provides the AM and FM demodulators along with AGC control signals, mute control and various base-band output lines. The module caters for both Narrow-Band and Wide-Band signals. The set-up procedure for this module can be frustrating since several of the links on the board, although clearly identified, their 'logic' is not marked (ON or OFF), like they are on the Transmitter Controller. Experimentation is the name of the game.