Back to Moto Test Equipment
Back to Home
  Repairing the Motorola R2001-series
Communications Systems Analyzers

By Steve Goggans K7LZJ
  Print this Page


There seems to be a dearth of repair information on the Motorola R2001 any suffix power supplies.

LV Power Supply part number 01-P07897V001 (A1) and HV power Supply 01-P07896V001 (A10) are used in R200x B and C series. NO relation to the D series.

This adventure started with the purchase of a smoked R2008C to get a replacement CRT for the dim fuzzy CRT in the R2002B I got at Dayton 5 or so years ago.

I just had to try fixing the C before I started pulling parts. The LVPS was bad and blowing line fuses.

The GREAT news is the R2002 CRT was just fine. After recapping and adjusting the LVPS and HVPS the old dim fuzzy CRT is BRIGHT and CRISP!

The "working" B LVPS was used for comparison and board swaps during troubleshooting. A lot of bad caps were discovered in the B supply by accident.

The low voltage power supply description says the switching power supply operates at 20 kHz. The operating frequency is set by R330 and C318 and both are fixed. The schematic value of R330 is 22K and is out of design range according to the MC 3420 data sheet value range of 5K to 20K. According to the data sheet nomogram, the values should be C318 0.002uF and R330 13K for about 20kHz - if you really care. AND what to my wondering eyes should appear on A1A3 boards in both supply boards R330 is a 13K metal film and C318 is 0.002uF. Correct your print. I tried substituting a 25K 20T pot for R330 and varied the switching frequency from 20kHz to 27kHz. If there was any difference it favored a higher frequency. Dead Time, Voltage and HVPS adjustments change with frequency. My take is I wasted my time!

First problem found was two out of three 2N6436 on A1A1 were blown: Q102 and Q104. Resistor R102, the base resistor for Q102, was cooked and is probably where the magic smoke came from. I tried many different audio and switching transistors to replace the unavailable 2N6436 before WA1MIK led me to the 2N6437, a higher voltage part that's otherwise identical to the 2N6436, and it IS available new (BOCA Semiconductor) on eBay. DON’T waste your time with anything else; get the BOCA 2N6437. While it seemed likely that almost any transistor would work at 20kHz switching frequency, they DON’T; in most cases they're not even close.

At this point the R2008 cards would work in the R2002 frame so the 2008 power supply motherboard was suspect. It was pulled and looked over. Nothing looked out of normal so back together again. It kept looking like PWM drive TP302 was not making it to A1A1 Q101 base. Finally I popped the power supply motherboard again and started measuring between connectors even though the traces all looked fine. The trace from A1-13 to A3-18 PWM DR was open even though it looked fine. Bright light and tipping the card revealed that the trace width under socket for A1A2 is extremely thin and was burned open. Also A1-14 to A3-19 Chopper DR B was intermittent. LESSON: If a power transistor on A1A1 board blows it may pop one or more traces under the A2 socket on the power supply motherboard. I added wires for the 4 traces on the back of board. Click on any photo for a larger image.

pix1.jpg

Dead Time Pot:

The Dead Time pot gets adjusted watching the waveform on Q103 and Q104 (see photo below) for just less than a sharp shoulder. Keep a close watch on the transistor temperature by laying a finger over them. Too far CW and they get HOT FAST! Back off the Dead Time pot CCW 5 to 10 degrees. On the scope trace, both channels are set to 5V/Div and the time base is set to 10uS/Div.

pix2.jpg

This is used to prevent the inverter switching transistors from cross-conducting at high duty cycles. In the repaired 2008 it was used to set the collector traces of Q103 and Q104 as mirror images while monitoring transistor case temperature with a finger.

In the R2002 supply it also makes a major difference in the CRT intensity and focus, so set it for the best display, Q103 and Q104 are also nice mirror images. The TO-3 cases run warm, not HOT.

Further experimenting finds the original Q103 and Q104 are quite forgiving for the setting of Dead Time.

Helpful Modification:

Board A1A3 "Dead Time" range resistor in front of pot R328 is R340: change 3.9K to 4.7K.

Test and Setup Notes:

The saw tooth on TP 205 reads twice the frequency on a Fluke 87. The 2002B frequency is 21.72 kHz on TP306 and TP307. The 2008C frequency is 22.39 kHz on TP306 and TP307. These frequencies are in the ballpark for the actual parts used on the boards.

Current limit pot: turn CCW till current limit starts toggling and then back CW a bit, maybe 15 degrees. Have experienced no problems and had accidental short with test lead in PS just caused the PS to cycle off and back on.

While the supply was out I recapped most of it and measured the removed caps. Some were open, some were leaky and some were OK. One had a broken lead. Well worth the effort. Ground leads are difficult to get out of the boards, as the solder on the topside foil is well heat-sinked. I used a Pace on the non-ground leads then a soldering iron on the ground one and pulled it out leaving the solder in the hole. Soldering iron on top of board and Pace on bottom cleaned them out for installation of new caps.

Along the way I found CR204, one of the Schottky diodes in the 5V supply, was almost totally shorted – easy to see on a Huntron Tracker and also the diode readings on a Fluke 87 were much different than a good one.

The PWM chip MC3420 seems to be a one off - no replacements. There are no other PWM control chips with dual outputs. The data sheet only addresses values of "Dead Time" values with a voltage on Pin 6 Vcontrol and in this application has Pin 6 grounded.

HV Supply:

High Voltage supply A10 has 3 electrolytics. The supply is easy to remove. Unplug the HV lead from the front of the CRT. Gently rock and pull the CRT socket. Remove the ground lead from the CRT center clamp. Remove the 5-pin plug from the scope deflection board A2 to the CRT socket.

pix3.jpg

Go to other side of chassis and unscrew the two captive screws on the bottom of the chassis by main chassis. It takes more time to read than do.

pix4.jpg

The card edge plug slides apart as you carefully pull the supply out and thread the plugs and CRT socket through chassis. Lube plugs and sockets with Nyogel - makes plugging in CRT socket a breeze.

R2008C caps were close and slightly leaky, R2002B C17 OK, C15 LEAKY orange wrap brown and PC board discolored, C21 open - bottom 1/4 inch of orange (photo looks yellow) plastic wrap is black. Note the red seal on the bottom of the bottom capacitor with the brown ring in the picture. This is one of the infamous "No Name" red sealed caps that you REALLY want to replace. Most of the "No Name" red bottom caps I found were clear plastic wrapped and say USA on the side.

pix5.jpg

The wires connecting to the CRT socket started breaking at the solder joint on one unit. Using an extender board is much easier if the wires are a few inches longer so some flexible wire extensions were added to solve both problems. The white ground wire on the socket had only a few strands unbroken and was replaced with a flexible black wire seen leaving the lower right corner of the picture.

pix6.jpg

General tips:

  1. The cards and connectors require a lot of effort to come apart / pull out. Nyogel 760G on the plug sockets and card edge connectors helps a lot. The card rails were tight and dry. A THIN coat (almost invisible) of THIN grease (I prefer silicone) on the card edge and slide brackets is a BIG help.
  2. Whoever manufactured this for Motorola didn't believe in heat sink grease. There was none in either supply. There was soot under the transistor mica insulators so a small amount of THIN silicon grease was wiped on the micas. Thin is the key word; too much is worse than none.
  3. TO-3 transistors on both LV and HVPS have nylon step washers on the board side. The step down that goes through the plate may be too long or have pressure flowed to be too long. These need to be trimmed to assure the TO-3 cases and mica insulators fit flush on the plate.
  4. Check the line fuse rating – 1.5A is correct; DO NOT increase the rating.
  5. If the BAT LED is on and no battery is hooked up - as was the case of both units, C307 is open. I used a 220uF, 50V cap as a replacement.
  6. A wire added to the collectors of Q103 and Q104 make handy test points while setting "DeadTime" to get good switching waveform. NOTE: Put a finger across both of them to monitor temperature. Too far CW and they get hot FAST. Go Back CCW a bit at a time until they will stay cool to barely warm for hours on end.
    pix7.jpg
  7. The Horizontal and Vertical size adjustments in the service notes can run out of adjustment. Tweak the 2KV setting pot on the HV supply to change their range. Get the HV off too much and focus will stay a bit blurry – letters and numbers will be hard to read. Turn the HV adjust pot CCW (higher voltage) on the HV supply and readjust intensity bias / limit on the scope card A2. Opposite if you need lower end.

Other Repairs:

Note: The front panel comes off with all the knobs and switches as an assembly and ribbon cables unplug from the main board. CRT center bracket needs to be loosened and tube socket removed to slide front forward. A few wires need to be unsoldered from the motherboard depending on how far you need to remove the panel. Use caution with the flex connector cables that damage easily.

Lubing the Step Attenuator:

While at it I decided to look at why the step attenuator was so hard to turn. The 2008C has a mechanical step as the buss interface is Cellular. The 2002 has a switched attenuator because it has the IEEE 488 interface.

The thick cover on the mechanical stepped attenuator with lots of little screws is sort of easy to get to. Avoid the temptation. There are tiny springs and plungers to hold the contacts solidly closed when not activated. Just in case you need a spring, I found one pulled out of a lock cylinder was a dead ringer. The lock was not a Kwikset, but I don't remember the brand. The Attenuator bracket screws can be removed with the input module swung out. The two SMA connectors unscrew and it drops out. The tin cover with 4 screws comes off and there is a cam shaft to actuate the switches. It looked like it had never been greased. A bit of thin grease on the cams and plungers made a world of difference.

Print Errata Not Noted Above:

Low Voltage Output A1A2 Schematic: L203 lead going to R203 has been pulled out of board and a 1N4001 added in series cathode to board. L204 lead going to C213 has been pulled out of board and a VSK520 Schottky added in series cathode to board.

Low Voltage Control A1A3 REV E Schematic: Pot R313 is 10K not 5K.

HV PS (A10) J1 pin 4 to LVPS J6 pin 3 goes to GND on LVPS MB. NOTE: HV PS J1 Pins 1 and 4 not tied together on HV PS.

Notes:

Nyogel 760G I was told by one of the factory product service engineers from a previous life, that it was developed by or for Nissan racing to solve intermittent connector problems. We used it on low-level signal connectors and card edge connectors in industrial control systems to make intermittent gremlins disappear. This is one of the products that lives up to its website hype. It was used on ALL the card edge connectors, plugs and sockets including the CRT – no more gremlins.

A few cross reference part numbers found along the way:

Test Points from two working units taken with a Fluke 87. Power switch ON and warmed up and TP202 adjusted for 5.2VDC.

TPR2002B Boards Recapped R2008C Boards recapped and Freq set to 22 kHz
TP10110.61 VDC, 6.94 VAC 43.5 kHz10.13 VDC, 6.5 VAC 44 kHz
TP10215.26 VDC, 0.76 VAC 21 kHz14.5 VDC, 0.87 VAC 19.5 kHz
TP10314.48 VDC, 0.79 VAC 21 kHz14.2 VDC, 0.88 VAC 20.5 kHz
TP20100
TP2025.2 VDC, 1.5 mVAC Adjust to 5.2V 5.21 VDC, 4 mvAC Adjust to 5.2V
TP203-4.97 VDC, 1.9 mVAC-4.91 VDC, 4 mvAC
TP20412.95 VDC, 1.8 mVAC13.15 VDC, 4 mvAC
TP205-13.00 VDC, 1.8 mVAC-13.40 VDC, 4 mvAC
TP20634.7 VDC, 2.7 mVAC34.4 VDC, 4 mvAC
TP207122.2 VDC, 15 mVAC117.9 VDC, 12 mvAC
TP208-121.8 VDC, 7 mVAC-117.9 VDC, 11mvAC
TP3018.02 VDC, 4.4 mVAC7.94 VDC, 4 mvAC
TP30215.64 VDC, 0.55 VAC14.95 VDC, 0.64 VAC
TP3035.03 VDC, 104 mVAC 38 kHz5.08 VDC, 100 mV 44 kHz
TP30412.81VDC, 73 mVAC 21.8 kHz12.77 VDC, 36 mV 44 kHz
TP3054.14 VDC, 1.08 VAC 43.6 kHz4.12 VDC, 1.07 V 44 kHz
TP3065.97 VDC, 4.8 VAC 21.8 kHz6.12 VDC, 4.54 V 22 kHz
TP3075.99 VDC, 4.8 VAC 21.8 kHz6.13 VDC, 4.54 V 22 kHz
TP40115.94 VDC, 0.51 VAC 120 Hz15.17 VDC, 0.61 VAC
TP40200
TP40315.32 VDC, 0.51 VAC 120Hz14.38 VDC, 0.603 VAC
TP40400.51 VDC, 55 mvAC - oven at temp for several days

Test information from attenuator: Motorola 58-P00476N001 / RLC Electronics Model A-2610D

Termination: Bird 50 Ohm measures 50.6 Ohms and includes test leads etc Fluke 87.

dB
setting
Input R
Out Open
Input R
Out 50 Term
Output R
In Open
In to Out R
0Inf53.2 Inf2.0
-1060.151.7 60.551.8
-2052.652.7 51.583.6
-3049.049.7 50.593.1
-4049.048.8 49.296.0
-5052.051.9 49.4100.2
-6049.349.1 49.597.9
-7050.750.7 49.7100.5
-8048.749.1 49.698.4
-9051.651.5 49.8101.4
-10049.049.0 49.698.5
-11048.848.9 49.798.4
-12048.648.8 49.898.3
-13051.551.5 49.7100.9

I only did one test run so there may be a mistake or two.

Contact Information:

The author can be contacted at: k7lzj [ at ] arrl [ dot ] net.

Back to the top of the page
Back to Moto Test Equipment
Back to Home


This web page created 04-Apr-2016

This web site, the information presented in and on its pages and in these modifications and conversions is © Copyrighted 1995 and (date of last update) by Kevin Custer W3KKC and multiple originating authors. All Rights Reserved, including that of paper and web publication elsewhere.