PDA

View Full Version : Not quite a boatanchor, but definitely in need of help



N8YX
04-27-2020, 09:51 PM
I recently fell into (or fell over) a JRC NRD-515 receiver. One might argue it's a poor man's version of the Collins 651-S1, though these rigs always have commanded a decent price.

Mine's in need of help. The knobs look to be coated with many years worth of grud buildup, the RF band pass filter selection circuit (and possibly several band pass filter elements) aren't working properly. On the plus side, it came with the accessory filter board and the OEM 600Hz CW filter. The negative to that is the CW filter appears to not work. It's a Kokusai and their failure mode is well known - as is the fix.

Walt, PE1ABR has been helpful with troubleshooting and providing reference voltage levels in a working unit (he owns several of them). First order of business is to find out what's up with the band pass circuit, then on to the filter. I managed to find a 1000Hz JRC filter for the unit so it'll go in the Aux filter slot.

I'm tempted to mod this thing with a bunch of the more widely known improvements and a couple of my own thrown in for good measure. Such as an IF notch filter and a squelch. Cosmetically, the receiver is a very close cousin to the FT-901/902DM (albeit with black instead of gray knobs) and that's what I'll probably end up pairing it with. I had a chance to buy a complete 515 station (including the transmitter and station power supply) but that setup is a little too pricey at the moment...

K4PIH
04-28-2020, 07:30 AM
Knob crud seems to be a universal thing with older gear. All Swan gear comes with knob crud. Some of it is just from crummy fingers but I also think the plastic used promotes mold growth. I take all the knobs and soak then in vinegar then use a tooth brush and antibacterial soap. I've also noticed that high humidity also promotes mold growth.

651-S1, used many in my AF years. Great radio, wish I had one.

N8YX
04-28-2020, 08:36 AM
651-S1, used many in my AF years. Great radio, wish I had one.

The one I really wish I had is the 451-S1 It's the receiver counterpart to the KWM-380/HF-380 and as I understand less than 10 were made. Would make for a nice pair of twins on the operating desk.

Back to the -515 momentarily. Appears that I am going to be firing up the desoldering station...

N8YX
05-08-2020, 01:31 PM
It's been a while since I started into the -515, so here's a blow-by-blow on what was found.

First, the radio appeared deaf on several bands (verified with a signal generator). The other apparent issue was the failure of the accessory CW filter (or was it? More on this later.) A dialog with Walt, PE1ABR regarding the working of IC61 (a uPB249 ROM on the Synthesizer Board) and its action on IC1 (an SN7445 8-line decoder on the RF Board which controls RF bandpass filter selection) motivated me to test IC1. Several of the decode lines appeared active (low) at once, while others never went low. Removal and replacement was in order.

N8YX
05-08-2020, 01:37 PM
JRC must have used the same PC board vendor as Kenwood did, as it's very easy to lift traces while vacuum desoldering. I kept the tool to 700 degrees F but apparently this was too much - one of IC1's pads was partially lifted. Chip-Quik is on the to-get list for the next time I need to desolder anything from the boards.

A socket was fitted, a new SN7445 plugged in, DC resistance checks performed and the RF Board re-installed in the radio. Checking with a signal generator revealed that all ranges showed a meter reading of S9 with 50uV RF applied to the antenna connector. Perfect.

16929

N8YX
05-08-2020, 01:55 PM
Next up was the installation of a 1KHz B/W filter onto the accessory filter board. (As mentioned in the first post, the radio had this board included along with the OEM option 600Hz CW filter.) I left the board out when performing initial RF bandpass filter circuit verification tests, though when I attempted to switch the radio's filter selector to the 2.4KHz spot the receiver went deaf.

Hmmm. That filter is present on the RF Board and even if faulty should pass SOME signal.

Re-installation of the board restored operation, though the 600Hz filter was way down in response. I started poking through the schematic, then looking at the board - both physically and with an ohmmeter. What was actually on the board - and the operation of the filter selector through its associated diode switching matrix - bore little resemblance to the schematic.

Time for an Ian Dury and the Blockheads quote: There Ain't Been Some Half-Clever Bastards. What this one apparently did was to remove some of the chokes present on the filter-select lines, install different types and reroute their outputs to other points on the schematic; this to allow the filter known as "FL4" (on the bottom of the board in the picture) to be the second filter in the selection menu. A Murata 3KHz filter was fitted at the FL4 spot when I received the unit and this had to be removed to allow installation of the 1KHz JRC crystal filter.

16931

Note the yellow arrow. That's actually the stock 2.4KHz B/W SSB filter, which was bad. The CW (600Hz) and RTTY (1000Hz) filters are on the board to the front of the SSB filter.

N8YX
05-08-2020, 02:00 PM
I never throw anything away, particularly those assemblies containing useful parts. A scrap TS-930S Signal Board yielded 3 47uH molded chokes; these were installed after the previous owner's handiwork was removed. The lower left arrow shows a stock JRC-spec choke while the upper-right part indicates the choke obtained from the '930 Signal Board:

16932

N8YX
05-08-2020, 02:13 PM
Once the filter switching was determined to be in correct order, failure of the SSB filter (and proper operation of the others) was confirmed. This necessitated another removal of the RF Board, removal of the filter then cleaning, restoration and replacement of everything.

Kokusai filters are known for breakdown of the damping foam used inside them. This link gives all the particulars and presents the reader with a set of instructions on how to recondition one: http://jlkolb.cts.com/site/koku.htm

Sure enough, upon disassembly of mine I found a gooey mess. Careful cleaning and re-packing with closed-cell shipping foam resulted in a working filter with the specified bandwidth characteristics.

The 3KHz Murata which was removed will be used in a separate circuit built onto an accessory board; this after the receiver has been brought to factory spec with an alignment of the internal signal generation stages. At present it's hearing well - even on a 10ft piece of wire in the lab.

N8YX
05-14-2020, 10:33 AM
Another R&R of the RF Board, this time to correct what may have killed the Bandpass Filter decoder IC:

Seems that JRC used two back-to-back pairs of diodes as a front-end protection circuit. Devices used were silicon; when you get more than 1.5V RF at the Antenna connector all sorts of interesting and undesirable intermod products appear in the receiver passband. Not uncommon at all if you live close to a MW broadcaster or run the radio in a multi-multi setup.

Counsel at the time was to remove these diodes. In the unit I have, someone did - and in doing so, possibly caused the 7445 bandpass filter decoder to be taken out when a nearby lightning stroke put a sufficient voltage pulse into the set.

I did a bit of research into receiver protectors and the various ways to implement one. Many still utilize a diode (or diodes) somewhere in the current path. One had an NE-2 bulb paralleled with a 1mH inductor; the inductor keeps the DC (static) potential of the antenna low while the neon bulb shorts a voltage spike to ground by way of firing. This didn't seem to be a viable solution for a couple of reasons - one of which being that the addition of the inductor pretty much killed sensitivity of the radio below 2MHz.

What I eventually settled on was a gas discharge tube in parallel with a bidirectional TransZorb. Think of a Zener diode with a very high amperage rating and the ability to conduct in either direction. The decoder IC specifies 30V max switched voltage at its outputs; the other devices (potentially) in the path of a pulse are the 1st mixers - a balanced pair of U310 JFETs. Their gate-to-source maximums are spec'd at -25V. A 13V TransZorb (right arrow) and a 75V Gas Discharge Tube (left arrow) were installed in the original vias. Should the TransZorb ever open after repeated conduction cycles, the GDT will still offer protection from large spikes - and is what Drake, Kenwood et al did in their various equipment lines. Additional measures including switching the feedlines to ground or disconnecting them altogether when storms are forecast are used, and the combination of all should keep the front end free of incident.

16934

N8YX
09-24-2020, 02:22 PM
Bumping this one up a bit.

I recently acquired another -515 and the matching NSD-515 transmitter. My original appeared deaf next to the new arrival, so I began an exhaustive process of troubleshooting. Involving replacement of both mixer JFETs (the U310s I'd mentioned).

We'll cut to the chase: Although the Transzorb (right arrow above) isn't supposed to be conductive at low levels of RF, it apparently is. Enough to knock 30dB or more off received signal levels. So...a modification of the original overload protection circuit is in order: Two pairs of four diodes in series (1N4001, 4002, etc) with the strings arranged back to back and wired in parallel with the suppressor tube. A 1 megohm, 1W metal-film resistor may be added in parallel to bleed off any static charges.

The second receiver has been modified by Rob Sherwood. He hooked me up with all the info and parts required to do this to the first one, so when I have the supplies in-hand I'll mod the first set and place it back in service.

I will say that a -515 is one of the quietest receivers I've ever listened to, and that doesn't imply a properly working one is deaf. No, not by a long shot.

W3WN
09-25-2020, 10:22 AM
Times like this, I sure wish you took side jobs.

K4PIH
09-25-2020, 12:36 PM
I like the GDT use.

N8YX
10-21-2020, 11:11 AM
Got Rob's kit of parts in last week and reworked the Signal Board. One of the additions was an IF buffer board in place of C291 - the AM detector coupling capacitor. This helps improve detector linearity on all modes and is especially useful if you're going to run an external detector for synchronous AM (SAM).

I plan to monitor the receiver IF outputs with a scope, so the boards will prevent loading the IF down when one is connected.

A few assorted eBay transactions yielded two Yaesu FRT-7700 preselectors plus some connectors. These were originally marketed for use with their FRG-7700 receiver and have two flying-lead pigtails (one for MW broadcast band and the other for SW) to connect to the parent receiver's push-terminal antenna jacks. Job One was to swap these out for an RCA male-terminated run of RG-174 (SW range) and a female-terminated version of same (BC range). Luckily, two of the preselectors side-by-side are exactly the width of an NRD-515 so I can put both on the 10-12M lineup's middle radio.

The reason for this...as with many dual-conversion receivers, the problem of images arises. Not a big deal if you're far away from high-powered emitters, but one of the 10M Gang is right down the road and occasionally goes QRO. A passive preselector will help reduce image interference.

So far, so good. When my order of RCA female connectors arrives I'm going to modify the NSD-NRD transceive cable and put a flying lead on the receiver side's RX Mute line so the secondary receiver can also be muted upon transmit.

N8YX
11-30-2020, 10:12 AM
Another update to the ongoing repair saga:

I'm a perfectionist when it comes to gear performance. If something isn't quite right I'll doggedly pursue it until the equipment in question is operating at its design spec. Receiver #2 (newest arrival in the lineup) worked as I'd expect one of these to operate. Prior to me acquiring the set, it was sent to Rob Sherwood for his modifications and a thorough alignment, Thus, I consider it to be the "baseline" against which the performance of any other NRD-515 should be compared - and this has been verified on the test bench.

Running the first and the Sherwood-modded receivers side by side, I noticed that the original was still a bit deaf. Chancing across an eBay auction for another one, I got it ("for parts or not working - No Returns") for not much more than the two accessory JRC filters (included in the unit) usually fetch. It was advertised as having very low sensitivity above 16MHz.

Never hurts to have another example to compare to, right? When I unpacked #3 and put it on the bench, I found it to be able to adequately hear a 2uV, 1KHz-modulated signal on the upper HF region but both of the Kokusai mechanical filters (2.4KHz/600Hz) included with the unit required repair. Once this was done I performed the Sherwood mods and installed one of his AM pre-detector buffer boards in the vias left by the removal of C291. Note that this is a well-documented modification, but since Rob had the boards on-hand I got a couple rather than simply rolling my own.

After all the work was completed, the Signal Board was returned to the receiver and a full alignment performed. This is where things get a little hinky as compared to the procedure listed in the service manual. One of the major complaints about this receiver is the absolute non-linearity of its S-meter. Another is that it isn't as sensitive as compared to, say, a TS-940S or later modern rig. If you follow the service manual, you probably won't get either aspect totally right. Through a bit of empirical tinkering and adjustment, I came up with a way to get the S-meter to track fairly accurately. The big surprise came yesterday and this morning after I'd put the newest acquisition in the place of the original. It hears things that #2 doesn't. This may be due to the addition of the buffer board, but I think more of it is due to the way the IF Gain adjustments were made.

Where does this leave #1 - the first receiver I'd acquired and restored? I applied the procedure to it and while the meter tracks accurately now, sensitivity is still "meh". I'd noticed the AGC motorboating when the set was first powered up with the RF Gain set at maximum - and I'm now of the opinion that the AGC subsystem is unstable, thus reducing sensitivity. I'm going to replace every electrolytic capacitor on the Signal Board as a starting point. #2 has had this done and is stable, though #3 appears to have the OEM capacitors on the board and works quite well.

I'll write the process up and take a few pics of #1's adjustment points when that repair is successful - just in case this bites another JRC owner. Walter, PE1ABR suggested a few things to check but most involved the synthesizer and VCO sections of the radio - and those circuits appear to be operating properly.

K4PIH
12-01-2020, 07:51 AM
When I desolder on sensitive boards, I use solder wick and just enough heat to make it draw.

N8YX
12-01-2020, 10:37 AM
When I desolder on sensitive boards, I use solder wick and just enough heat to make it draw.
You need a vacuum desoldering station to take the components off anything JRC built. The vias are plated through and eyelets are present on both sides of the board. No more that 670 degrees tip temperature, either - or you'll start to lift the foils on the solder side.

A DIP component gets into a whole 'nother level of frustration here. I highly advise cutting all the pins, removing one at a time then installing a high quality socket.

N8YX
12-03-2020, 09:55 AM
Panoramic view of the triplets. This is the transmitter, #3 receiver (center) and #2 receiver (right). The center unit is the most recent repair. Parts coming in for #1 receiver and it'll be used as a test fixture for some other mods I have cooked up. The FRT-7700s on top of the center receiver feed each one, respectively. A local CB channel and Radio Havana were being monitored concurrently, while the transmitter was warming up awaiting that night's ragchew session on our 10M hangout.

17320