BRC/Thorn 3500 Chassis
Model: Ultra 6714
Original List Price : £255
CRT: Mazda A56 120X
The set has now arrived, its a “Mice” example ( that’s intentional, so read on) of the Ultra branded 3500 from Thorn. The sets cabinet has unfortunately seen better days, the bottom left has been gnawed at by mice. They have then charmingly puked the sawdust up, inside the cabinet ( see photo below). The damage to the cabinet although annoying and a shame, should be fairly straight forward to repair with filler.
Upon quick inspection the chassis appears to be in pretty good shape and the mice seem to have left the boards alone. The power supply module looks rough but compared to the recent HMV 2703 module, eminently fixable. So this is another missing CTV manufacturer added to the collection, just need to find and EKCO and a Murphy….a…….
First Lets Check The CRT
I wanted to see what the A56 120X is like and OH dear! what a disappointment. First of all I could hardly set the Cut-Off for the red gun, only managing to raise it to the required +1 division by having the G2 control at maximum rotation. Green and Blue Cut-off’s were set normally. When emission tested well what can I say other than its pants!
I left it running on the B&K for an hour and then rechecked Reds Cut-Off, there was a slight improvement but very minor. The emission did not improve whatsoever. Next I tried clean and balance, this restored Red and blue but Green was having none of it.
So the only option open to me now was a rejuvenate of the Green gun. This achieved the desired result giving Green the lifty that was needed.
Well at least the repair is now viable and I should get something decent from the A56 120X.
Next Stage:The PSU module loaded on to the test rig for first power application and based line evaluation
The Power Supply
The Ultra’s PSU module outwardly looks fairly OK, the most obvious first problem to address before power up is C607, this is in the 30V supply, no 30V rail means no ‘Chopper’ drive therefore no ‘EHT’, no ‘Field scan’ or ‘sound’ So the three rails 240V, 30V and 58-65V will all need to be checked out and to do this rather than using the set I use my Thorn test rig.
I guess on the 3000/3500 the single most troublesome module is the PSU, I’ve had no end of fun with these and having repaired four of them so far this year, I’m starting to get to grips with them as you can follow over here. In fact I’ve spent so much time working on Thorn PSU modules I have a dedicated work bench with the test rig, relevant data plastered on the back wall and parts all to hand.
During the life cycle of the 3000 series chassis, C607 underwent a number of value changes. I’ve found 1000uF 50V, 1250uF 63V and 1500uF 63V in this position on various PSU modules. I would hazard a guess this one was a hurried service call to a customers house, as the connections have another snipped caps tabs tacked to the the in situ.
The one fitted here is 1000uF 63V and it has to be the worst one I’ve ever come across. Just look at the poor thing, don’t you just feel sorry for it? You can just see it had enough.Out of interest and perverse pleasure I wanted to see just how sickly it was, so on to the torture table reformer. I was quite surprised and had huge admiration for the little thing, it happily held 50V without any sign of leakage. Hang on though the cap is rated at 63V and I normally always go 10-15% above. On to the next range and the little old cap soldiered on but above 50V and just shy of 70V, it couldn’t help itself and incontinence set in. All just a bit of fun!
Another casualty being the Dynamic trip X601, a Mullard VA1040. So to get to X601 to replace I’m going to have to do the usual spatch-cocking of the PSU module and not forgetting to replace C607.
With that all done I hooked up all the extension leads in the rig and powered up. Observing the rail indicators 6.3V came up, 30V came up and the 58-65V appeared. This all happened in about 1.5 seconds before the rigs trip operated ( same as the type in the 3k set) with a puff of the magic smoke appearing from R609.
With R609 (15R) overheating we have many possibilities and I need to check them all starting with the Chopper transistor (VT604) as it could be short-circuit. Then I need to check W606,W609, W616, W620 for open-circuit. There’s the Crowbar SCR W621 and the 72V zener W617 either of these could be short-circuit. The Feedback amplifier VT608 could be faulty causing the mark-space ratio of the monostable to allow the 58V to 65V rail to increase above the strike voltage of W617. Not forgetting the possibility of the Mica washer on chopper transistor having failed.
Stating with the Chopper transistor R2010 (VT604) that was found to be short. New R2010 installed, new mica washer and heat-sink compound. All PSU rails are now working. Despite now repaired whilst the module is opened up, I’m going to take the opportunity to replace a few of the tired caps and generally tidy up the few tacked in diodes. I will also run the scope over it to check the waveforms against spec. The I will hook up the line timebase module to the rig and PSU and again check the LTB waveforms.
Now to check out the waveforms on the Ultra’s PSU. Happy so see the 3500 switched mode PSU is performing faultlessly when various sections such as the Mark space, chopper etc were scoped.
Next Stage: The line time base module will be added to the PSU on the rig and tested. Once I’m happy these two modules are functioning then they will then go back into the set to see what we get on screen and how the I/F, Video, Decoder, frame and sound stages are fairing, hopefully something on screen from the CRT.
Line Timebase Module testing
Connected the LTB up to the PSU on the Test rig. First thing to do is remove the fly-lead that has been attached to -ve end of W608. This is used to simulate the line pulse so you can check the 58-65V rail without having a LTB panel connected. Now we are connecting a LTB it must be removed.
With the Line Timebase connected and powered via PLG on the PSU, it was time to power on. All the rails came up and with an insulated screwdriver I was able to confirm the EHT pulse output was present on the EHT transformer nipple. Finally I scoped the chopper pulse and oscillator output, all looking very good. Chopper pulse and oscillator output there albeit at the wrong frequency.
I reinstalled the LTB and PSU fed in test card F then powered up to see what I had on screen. What I got was a complete Red Raster with the line frequency way off. Adjusting the horizontal hold would not pull it in. I then shorted the flywheel to chassis and tried adjusting L501 to get a float, this was not happening either. OK looks like there is an oscillator fault that needs further investigation.
The Line Timebase module was removed from the set and back to the bench. C502 & C506 were innocent. As a good measure I also removed and tested R524 ( 250R) horiz pre-set, it was OK. Removed the Beam limiter board to reveal the underside of the TMB panel this revealed two faults. R528 ( 18R) had exploded, all that was left is a little stubby end. There was a loose wire is the from the junction of W904/R907 ( beam limiter) connecting to the line output stage earth.
I replaced the flyback tuning cap as the green meanie was not to be trusted. C502+C506 were fine and as a precaution I had replaced C511 as it was on of those little orange jobbies that don’t fair too well. Process of elimination, removed and tested VT501 it tested OK. Removed W501, W502, they test OK. Next W503, this is when I find one side of R527 goes to ground the other side should be soldered to W503’s cathode which in turn connects both to the oscillator ( VT502) base. However the tweezers reveals that end of R527 it is not at all its just resting ever so close but not connected. What’s more when R527 is tested its also out of tolerance. It should be an 82K 10% so max being 90.2K it reads 97.9K but regardless of that fact, not being in circuit is the likely culprit, well fingers crossed. I tested W503 for good measure and it OK
To prove I’m not mad and the jig is not fibbing and working correctly, I reinstall all back into the set. I then try to set up R504 by the book or in the case the service manual. I advanced the brightness to maximum. Shorted the flywheel TP to chassis. Using an AVO 9 ( Ruggerdised 8) and connected to the slider of R504. The objective being to set the voltage via rotating R504 until it reads 6.2V. Then adjust L501 (osc coil) for a floating picture, remove flywheel short and there you have a lock. The voltage at one end of R504 is 17V, at the other end 12V, so unable to get down to 6.2V
Back to the jig, bench and scope to try and find what must be blindly obvious but is eluding me thus far. I’d completely over looked the blighter that was causing all this trouble C508 an decoupling electrolytic connected to R504. It was Out of value!!!! High ESR!!!!! and was electrically leaky!!!! well that’s really not going to help is it!!!
It is now working, R504/L501 adjustment, as you can see period of 64us and Freq of 15.625kHz Back into the set tomorrow and tackle perform R504 set-up, then tackle the video fault which is likely the thick film.
We have lock, It looks like the decoder is working as I have colour, that’s a plus. First thing I’m going to do is replace the ‘Thick film’ on the video board to see if that clears up the all red problem.
Removed the video board and gave it a good clean. Found a NOS ‘Thick-Film’ unit in stock so soldered it in and tested. The red overkill was gone replaced with a ghastly mess, I would hazard a guess as this thing was dying the RGB A1’s were adjusted to compensate. Did a rough and I mean rough twiddle of the A1’s to give a reasonable picture. Not looking too bad, needs grey-scaling and some convergence. I will go through all the set-ups on the video.
I finished removing, cleaning and testing the remaining boards namely the decoder, Frame/Sound and I/F panels. Running the ESR meter over the all electrolytics, for the most part they showed to be within tolerance, some borderline and will no doubt fail but that’s OK I prefer that and do fix-on-fail when they do. I did a final grey-scale and convergence moving any errors as such out to the edges as you’re supposed to.
For a fairly ratty ( literally) high hour set which had lets be fair a knackered CRT, now rejuvenated its giving the best account it probably can. The body count of replaced components to get to this stage is relatively low, as I said above more may fail with further use but for now that’s it.