AR2002 bulletin page

The AR2002 receiver is no-longer serviced. Parts are not available from the manufacturer and service costs equal or exceed the resale value.

After several years of use some of the LCD display segments can become faint. This is not due to a failing LCD as a lot of people think but is simply caused by poor connection between the LCD and PCB. To cure the problem, both cases halves and the front panel will have to be removed from the set. The case halves present no problem (6 screws) but care has to be taken with the 4 countersunk screws holding on the front panel - the plastic around theses has tended to become brittle with
age and cracks or broken lugs are common. The easiest way to get at the display is to remove the CPU board completely from the set - usually 4 screws and 2 connectors. Remove the 2 further screws holding the sub board to the CPU
board and hinge it out of the way. To remove the LCD, remove the 5 screws holding the metal frame in place. The display should now free but remember where the insulation pieces go before removing it and note the order that all pieces are fitted.
To clear the fault, carefully remove the contact zebra strip and clean it (alcohol, IPA, etc).
Clean the PCB contact pads with a fine fibre brush or some very fine abrasive paper.
On rebuilding the set all segments should now be bright and clear.

*While the front of the set is stripped down, it is a good idea to check the display lamp.

If the unit crashes for any reason (supply spike or whatever), the AR2002 can be reset in the same way as the AR2001 (see AR2001 RESET for details). All details are the same including pin numbers. One point worth noting is that two different types of back up capacitor have been used in the AR2002. Earlier sets used a 1 Farad item and later ones tended to be fitted with a 0.47 F capacitor. The expected memory retention time from a fully charged unit (ie. one that has been connected to the power for a few hours) is approximately 5 days and 3 days respectively.

The AR2002 has proved to be a very reliable and stable set. As with all crystal referenced sets, the frequency may shift to some extent with time. If this has occurred then two points need to be looked at. Note that any alignment should be carried out at full working temperature.
First of all the PLL REF xtal needs to be aligned. IF this is out, any further alignment will correct the set at a spot frequency but will probably push it further out on other frequencies. An accurate frequency counter is needed to do this. To align this, either sniff the frequency directly from the xtal (X1 on PLL unit) or take from pin 17 on the PLL IC upD2833C (IC7 on early sets or IC5 on later). Take care not to load or pull the frequency by direct connection. The frequency should be 6.4MHz although some units may be fitted with a 3.2MHz xtal. Once this has been aligned with TC1 the rest of the alignment can be done by ear. On a known VHF frequency, align T2 on the main unit for best signal. This is located in the
top front area of the set surrounded by a metal shield near to the 46.998MHz xtal. In a few cases T10 & T11 may be peaked for best signal.

The AR2002 is normally a very sensitive receiver and as such can be damaged by exposure to very large RF fields, static or lightening through the antenna socket. Damage is usually limited to failure of the protection diodes and RF amps. In most cases only the 25-550MHz stage is damaged but occasionally the 800-1300MHz front end may be at fault (if it was in use at the time). The components to check are D18 (1SS97) and IC1 (MC5800) if 25-550MHz is insensitive and D19 (1SS97) and IC9 (MC5805) if 800-1300MHz is at fault. All of these items can be found on the main (top) board in the rear right hand corner.

The line has a range of 4.9V at no signal to 0.7V with a large signal. AGC level is driven from Pin-11 of IC3 (NIS112) via Q36. To disable the AGC, cut either R127 (33k base feed to Q36) or D12 (collector supply to Q36). A false level will now have to be fed onto the AGC line to give approx 4.9V, either feed a 5V level into the AGC live (derived from 6V supply on Pin-1, J3 main unit) or feed a high level (5V?) to base of Q36 via a 33k resistor (assuming R127 is cut) to give a constant 4.9V on the AGC line. The AGC line can be looked at on Pin-3, J4 remote connector PCB.

The remote control port is NOT an RS232 but just access to the CPU, it requires all the clocking information to drive the receiver.

Originally the RC-Pack GBP 255 STERLING was available to control the AR2002 but has not been available for many years. Another interface distributed by Garex (Scanmaster1- in the UK) was also available many years ago.

We have a rough faxed copy of the RC-PACK, this comprises of three pages of circuit / schematic and command list. The information is presented as a small PDF file (78kB), click here. No level details etc are available and we cannot technically support you should you wish to build one up. Most parts for the receiver itself are available from stock.

A scanned copy of the 15 page RC-PACK instructions is available here as an Acrobat PDF file (1476kB), click here.

For pager (POCSAG) decoding, it may be necessary to derive a discriminator (detector) output directly from the IC. A single wire may be soldered directly to IC4 MC3357 pin 9 and terminated in a suitable connector. The earth may be taken from any suitable point or from pin 15 (GND) of the same device. There are two MC3357 employed in the AR2002, the second being IC5 for WFM.

Both the AR2001 and AR2002 use a large ‘SUPER capacitor’ for memory retention and clock when power is removed from the radio. If the radio is left permanently connected to 12V (even with the radio switched off), power is fed to the microprocessor backup circuit so that memories & clock are maintained indefinitely.

The operating manual optimistically suggested that the memories would be maintained for up to seven days, in reality, we find up to four days more typical (3 days when the 0.47F is used and 5 days where the 1F is used).

However, memory loss may not just be due to a failed capacitor, there can be other reasons… here is a possible list:

To eliminate the likelihood of a failed microprocessor, it is a good idea to measure the backup voltage and current supplied by the capacitor when power is removed. The microprocessor takes around 5uA (five micro-amps) so measurement can be difficult.

If you don’t have a suitable digital meter, perhaps it is worth considering connecting a 100k or 1M OHM resistor in series with the positive terminal of the capacitor then measure the voltage drop across the resistor… the results may not be perfect due to ‘loading’ but should provide some indication of whether a problem exists.

For example, if the equation I = V/R is used (I is current in Amps, V is Voltage and R is resistance in OHMS), a voltage drop across the resistor should be in the region of 0.1V to 1V, if higher, the microprocessor must be considered suspect. However changing the microprocessor to fix this problem may not be considered cost effective… you may wish to live with the inconvenience.

The value of the SUPER capacitor can be either 0.47Farad or 1.0Farad. The 1.0F item is physically larger than the 0.47F capacitor. Re-charging a completely flat back up capacitor can take a couple of hours, the capacitor will re-charge whenever power is connected to the set, even if the set is not switched on. A fully charged capacitor should retain memory and clock information for up to 3 days for the 0.47F and up to 5 days for the 1.0F items.

It is unusual for the capacitor to fail but remains the most likely cause if memory retention time has dropped.

Although it is recommended that the capacitor is replaced with the same type as already fitted, it is possible to replace a 0.47F capacitor with the 1.0F item. The leg spacing is different between the two capacitors, therefore, careful bending of these is required to make the larger item fit the PCB. If this is done correctly, there should be enough space on the board to fit the physically larger capacitor in place.

Remove the bottom cover (4 screws) and unplug the speaker. Take care not to strain the speaker wires.

Locate the back up capacitor on the PCB now exposed. Take note of the polarity in which it is fitted.

Remove the board from the set (4 or 6 screws depending on type of board fitted).

Note: On the earlier AR2001 and possibly the odd AR2002, the shield at the rear of the set and two earth spring plates will have to be removed first (4 screws on side, one board screw and a piece of conductive tape).

Note: For all soldering required, quite a high wattage soldering iron is required (50W or so is ideal) – the earth connections tend to drain away the heat very quickly.

Fit the new item and solder in place taking care that correct polarity is observed.

Re-fit the board and plug in the connectors – these will only fit one way round.

Connect power to the set and switch it on. As all the connectors have been removed, the set will now perform a full re-set (displaying all LCD segments) and start working after a few seconds.

Replace the bottom cover once you have checked that the set is operating normally.

Leave the set connected to the power for a few hours to get the charge level topped up on the new capacitor.

There is no specific method for s-meter set up in the AR2002 service manual, so the best advice from the UK workshop is given below.

If the adjustment has never been touched, leave it alone. If it reads inaccurately, something else is likely to be faulty within the receiver (lack of IF gain etc):

The AR2002 was originally set up to give the following indication (NFM @ 145MHz):

However, due to batch variations and tolerances (different manufacturers of transistor in the IF stages), these reading are incorrect. VR1 & VR2 interact with each other so set up is very much trial and error. Considering their age, they are now likely to be quite fragile and intermittent.

Looking at the circuit, it is suggested that you try setting up VR2 to give the meter start point with no signal, adjusting VR1 for the first RED LED at whatever signal strength you prefer. Repeat this process until its correct (in your view). Remember to check that all LEDs go out when no signal is present.

The calibration of s-meter changes quite a bit with temperature, so you may need to leave the radio on for half an hour or so before starting adjustment.

Due to continuous developments, AOR reserves the right to make design and specification changes for product improvement without prior notice. The performance specification figures indicated are nominal values of production units. There may be some deviation from these values in individual units.

AR2002 Bulletin Page