A couple of weeks ago for my BBC Model B, I managed to pick up a non-working Microvitec 1451MS and due to this at a greatly reduced price. That was successfully repaired, the fault turning out to be a failed IC301 TDA1170, it now sits happily with my model B.
Since then, a kind donation from @hurty means I've acquired a Softel SE3010 Teletext editor terminal, basically a BBC Master. It needs a monitor, I really wanted to find another Microvitec to save having to keep swapping the one over from the model B. As luck would have it, another non-working one turned up, again at a bargain priced, today it arrived.
I wasn't sure what model it was but as so as I put it on the bench close to the 1451MS, without looking at the back I could tell it was 1431MS. The higher resolution Microvitec's used a darker phosphor, this you can clearly see in the photo below, the 1451MS on the Model B has the darker phosphor.
Just to reiterate, the model differences are as follows:-
1441 & 1442 both with a resolution of 895(H)x585(V), DOT PITCH 0.31mm
1451 resolution of 652(H)x585(V), DOT PITCH 0.43mm
1431 & 1432 resolution of 452(H)x585(V), DOT PITCH 0.64mm
2030 Resolution of 505(H)x585(V) elements, 80 characters on 6 dot, wide matrix, DOT PITCH 0.8mm
2040 Resolution of 940(H)x705(V), DOT PITCH 0.47mm
What other differences are there between the 1451 & 1431 monitors
Besides the resolution, the main difference of the 1431 is in the switched mode power supply. Being the earliest version of the monitor, Microvitec used for the chopper transistor, a BUW81A, which is what this monitor has. Back in the day, this transistor became hard to get hold of, so Microvitec modified the circuit to use either a TIPL753A or an R3213.
My other monitor, the 1451MS chopper transistor, is the R3213. However, this also required a different chopper transformer to be used, so they are not interchangeable. There were also other component value changes made to the PSU on the 1451.
1431
R8=56K .5W
R11=47R
R16=3.3K 5W
C10=22nF 400V
L2=4.7uH
Chopper transistor BUW81A
Chopper TX=PC3769
1451
R8=3.3K 1W
R11=22R
R16=470R 11W
C10=56nF 400V
L2=10uH
Chopper transistor TIPL753A or R3213
Chopper TX=PC5287 or PC5307
What do I know about the fault with this monitor
The vendor stated the following:-
It was dead and I found the PSU had blown up. I replaced both BR103 Thyristors and the BUW81A Darlington as well as a small handful of components and it was up and running for about 4 hours when it decided to die again, the BUW81A had gone short again taking out the 1R5 2 watt resistor in the emitter circuit. At this point a rang the customer who didn't want me to continue (the transistors are about £15 each!) and told me to sell it on
I checked my stock and no BUW81A in stock. I checked my books, and it would appear a BUT16 is an equivalent, I don't have any of those either. I can't consider using either A TIPL753 or R3213 as I would need a replacement chopper transformer, so I have to find a BUW81A or BUT16. Looking online they are indeed very expensive, however I found some new old stock BUW81A in Greece, so ordered three, they won't be here until early December.
This gives me time to study this early incarnation of the switched mode power supply and consider what is likely killing the chopper transistor, and come up with a plan to not deplete the only three transistors I will have. The most curious part of the fault description the chap gave, was that it was OK for 4-hours!
CrustyTV Television Shop: Take a virtual tour
Crusty's TV/VCR Collection: View my collection
The most curious part of the fault description the chap gave was that it was OK for 4-hours!Just being run out of its SOA?
In which case I would have hoped the over voltage over current would have stepped in. Perhaps that's a good place to start?
CrustyTV Television Shop: Take a virtual tour
Crusty's TV/VCR Collection: View my collection
It's as good a place as any...
Do you have a x100 scope probe, Chris? I'm curious to know what's going on at the collector of the BUW81A. If you recall, many years ago I repaired a Tektronix TSG271 which had a PSU fault and it was only by scoping it that I learned a resistor in the snubber network across said transistor had failed, leading to spikes that could cause failure under certain conditions.
Checking the emitter current won't be a problem here.
Snubbers are critical in these circuits, don't replace with next available value, it must be spot-on. At 800 volts there's not a lot of Vceo to spare.
A suitable power mosfet or a BU126 / BU326 or similar would work there.
Do you have a x100 scope probe, Chris?
Short answer is no.
Snubbers are critical in these circuits, don't replace with next available value, it must be spot-on. At 800 volts there's not a lot of Vceo to spare.
A suitable power mosfet or a BU126 / BU326 or similar would work there.
Not having a background like yourself in the electronics industry, perhaps I can be forgiven for not understanding how this is possible.
I will explain why I have difficulty. It seems Microvitec went to great lengths to match the chopper transformer to the chopper transistor, and the surrounding component values. The two types used were not interchangeable, each requiring mods to do so. You even admit yourself with regard to snubbers, values are critical. Yet, I'm supposed to take a leap of faith that a BU126/BU326 are drop-in.
I'm not challenging you or disagreeing, just demonstrating why I fail to grasp how this is possible. As always, happy to be educated.
However, rather than going down a rabbit hole of modification, I will be endeavouring to find out why the BUW81A gets destroyed, starting with the 'OVOC' (over volt, over current) circuits.
CrustyTV Television Shop: Take a virtual tour
Crusty's TV/VCR Collection: View my collection
The snubbers and the transformer do very much need to be worked out in conjunction with each other. The transistors are (or should be!) switches and so do not play as critical a role.
The BUW was an early attempt to improve turn-off time by not saturating the switch. The later BU126 addressed this issue in other ways.
Given that the original device is a Darlington and the suggested devices aren't, I'd ignore the suggestion and stick with the original device. And as you say, much effort was made by the folks in Bradford to make the PSU reliable.
Regarding the 'OVOC' circuit, by all means check it - it's good practice, after all - but do heed my trials with that Tek TSG!
BUW81A that is a transistor I haven't heard of for a while. I think they where used in Link 110 camera power supplies? There where two in series and you had to match them, by hand looking at the switch crossover point when under load.
Now there are no real good data sheets that I can find on your device, they don't mention it is a Darlington device, but the diagram shows this is and as it has an HFE (current gain) of 200. That is a lot of gain for a power switching device of this era, they are normally between 10 and 50 so this proves it to be a Darlington device. The BU326 is a good device but will not drop in here, just does not have the forward current gain required.
The two things that destroy power switching transistors are a) not enough base drive and b) to high collector/emitter or collector base volts. I agree with most check the snubber components C16, R6, D7 not forgetting C17, R12 and diode D17. Replace the two electrolytic capacitors in there C12 and C23 I would change the overcurrent C18 too. Lack of current driving the base of the transistor is difficult to see here as it is driven directly of its own winding.
When you fit the new transistor run for a couple of minuets then switch off and check if it is running hot. If it is it suggests lack of base drive. You do need a X100 probe to look at TP3. I hope they give you the waveform here. Excessive volts here normally kills the transistor almost at switch on. Welcome Crusty to the world of switch mode power supplies, yes in its day it stressed many of us!!
A Tandberg CVC2 I think it was, had destroyed the BU326 went it came in to me. Checked all components around the device but didn't see anything wrong. In with a new BU326 and yes it was destroyed almost immediately! very embarrassing to me! what to do next? I needed a variac transformer and also a low voltage isolated supply to get the switching driver circuit working. Did this to see at about 50V ac input that 100Hz mains hum was present on the collector of the BU326. The 100uF 400V mains reservoir capacitor was open circuit, wasn't expecting that. but solved. Why do I get all the bad ones! Its because you don't give up and have to know why? and fully understand what was going on and not let anything beat you.
. I think they where used in Link 110 camera power supplies?
No, pretty sure they were BU104s - the "suicide pact" transistors. The 120/125 used BUX80s.
I'm going to lay odds on the snubber components across the primary: that bl**dy 270k resistor in Tek SPGs/TSGs has now become a "change on sight" component.
Although I'm not from an electronics background and thus not really entitled to pass comment on a so-called professional repair, I am going to stick my neck out and do so. I'm not impressed at all, thus far I've found 4 power resistors with dry joints. I also checked the snubber component, R6 & D7 which are OK, however C16 should be a 2200pF (2N2) someone has fitted a .01 (10nF) so far out for what is a "critical value".
I think we may have found the chopper killer.
C224 is still there, which is another change on sight part. This entire board needs a thorough servicing, which it will now get, whilst I await the arrival of the BUW81A transistors from Greece.
p.s.
x100 2kV scope probe on order
CrustyTV Television Shop: Take a virtual tour
Crusty's TV/VCR Collection: View my collection
Although I'm not from an electronics background and thus not really entitled to pass comment on a so-called professional repair, I am going to stick my neck out and do so. I'm not impressed at all, thus far I've found 4 power resistors with dry joints.
You’re learning! It imperative to check these components and anything that run hot. Sometimes you may need to remove the components fully and clean the legs to get them to take solder properly. Also with any pulse caps, although the soldering can look ok check the capacitor body for any signs of distortion at the legs.
That polystyrene capacitor also jumped out to me when I saw the pictures, they were often troublesome. You might also want to give L2 a thorough test but as the power transistor is a Darlington it may have offered a little more protection to the coil than being a standard transistor going c-b-e short.
John.
Seconding John's comments above.
Just a small tip: if any capacitor is subject to pulse/fast rise time waveforms (such as those in snubber networks), choose one with a polypropylene dielectric. They're designed for it: polyester ones aren't.
There was talk on one of the BBC facebook groups about eight months ago about a batch of fake BUW81A from Aliexpress, could be that yours had one of these fitted. The ones identified as fake were opened and looked almost identical inside to the fake 2N3055 transistors also floating around.
Some 15 years back I learnt a lot about snubbers while modelling a POE (Power Over Ethernet) circuit in LTSpice.
The appnote circuit had no snubber, which I though was odd. When I simulated it I had a 1kV spike on the drain of the switching FET, way over its rating!
A simple R-C snubber fixed that, but I was surprised back them just how critical the values were. Get it right and the spike is completely removed without wasting a lot of power. Magic!
Once the POE circuit was on the PCB I ran the scope over it and found the simulation was very close to the real hardware. But your models need to be accurate to get this close prediction. That's where the peer review of the on-line community helps.
Oh BTW, for those who've never used it, LTSpice is a free download and is well supported by a large user community. Once you get used to the quirky schematic drafting UI it's very useful. Just the thing for trying out that circuit idea you've been going to breadboard and for testing "what if" scenarios in existing designs = you can see just how close to not working many TV circuits are!
Even if you just use it as an alternative to hand-drawing circuits it's well worth downloading.
Agreed re: LTSpice - have used it for some time. It's pretty quirky software full stop (it does betray what I suspect are Unix origins) but very handy to have in your arsenal. Linear's filter designer is another useful tool.
@freya it's unlikely (I would hope) in the case of the devices concerned but some power transistors use beryllia as an insulator and you really do NOT want to be cutting those open.
I recall a SMPS from EMI Oz in their first colour TV that used a beryllium oxide washer under the BU126.....
Beryllium oxide was commonly used in heatsink paste.
And what was that nasty stuff that was used in power factor correction capacitors?
Some 20 years prior, chassis were commonly plated with cadmium and asbestos was used as heat shields.
I could go on! RoHS be damned!!!
And what was that nasty stuff that was used in power factor correction capacitors?
Polychlorinated Biphenyls, or PCBs for short.
See also Toluene Diisocyanate, released from self-fluxing enamelled copper wire.
As for asbestos - see every Bush DAC90
