|
Vintage Test Gear and Workshop Equipment For discussions about vintage test gear and workshop equipment such as coil winders. |
|
Thread Tools |
12th Nov 2022, 1:20 pm | #61 | ||
Octode
Join Date: Jul 2009
Location: Cheshire, UK.
Posts: 1,733
|
Re: HP 5300B Overcurrent
Quote:
To get voltage on C6 requires contacts on the 50 pin connector bridging out & the power switch to be on, or as I did last weekend, linked with the orange clip lead (bypassing the switch too). Quote:
David |
||
12th Nov 2022, 1:57 pm | #62 |
Nonode
Join Date: May 2016
Location: Cornwall, UK.
Posts: 2,333
|
Re: HP 5300B Overcurrent
Yes David - I was referring to the flowchart type tests where power is applied from an external PSU across C6. With the mains tests on the variac, pins 25-50 were bridged and the power switch was in circuit.
PJL - no, the zener has never failed while testing with the PSU across C6, up to frsimen's 22V maximum. The 'scope is nominally 15MHz. Twiddling the trigger made no difference to the duality of the trace, but I can solder in another CR10 and try to provide a more spaced out waveform picture. I have a rather more advanced Tektronix but it's in storage as I'm living in a half-built house and keeping test gear to a robust minimum... Thank you all again for the heavy lifting you're doing in the brain department. This complexity of machine is well beyond my current level of being able to troubleshoot alone! |
12th Nov 2022, 5:15 pm | #63 |
Dekatron
Join Date: Sep 2005
Location: Seaford, East Sussex, UK.
Posts: 5,997
|
Re: HP 5300B Overcurrent
Replace the Zener, monitor Q17 collector with your scope, and slowly raise the supply voltage to about 28V (David's unit is running at 26.5V so it should be OK). Watch the trace and back the supply off if it starts misbehaving and take a picture of the trace at around 26.5V.
The chirping on mains sounds like it could be a stability issue. |
12th Nov 2022, 11:04 pm | #64 |
Dekatron
Join Date: Sep 2005
Location: Seaford, East Sussex, UK.
Posts: 5,997
|
Re: HP 5300B Overcurrent
Regarding the fault finding chart, the 10V pk-to-pk does not agree with the waveforms of figure 4-6.
I now have a better understanding of how it works having carefully read the explanation in the service manual. The on-time regulation relies on secondary voltages so I assume that it free runs until these voltages have materialised driving A1T2 into saturation which would cause A1T1 to stop driving Q17. Post #12 shows a frequency of just 8KHz when the supply was just 6V although there was a fault on -5V at the time. |
13th Nov 2022, 11:05 pm | #65 |
Octode
Join Date: Jul 2009
Location: Cheshire, UK.
Posts: 1,733
|
Re: HP 5300B Overcurrent
Done some checks on mine, it also chirped with less than 8V supply, but tried again later and it stopped chirping & managed to test at 6V, wonder whether a capacitor was reforming?
Mains voltage was 242V today. First set of tests of Q17 collector waveform with no module fitted. 8V DC input, 0.43A, 44V peak, 28V pulse, 28us width, 10.16kHz. 12V DC input, 0.27A, 47V peak, 30V pulse, 24us width, 16.23kHz. 15V DC input, 0.21A, 48V peak, 34V pulse, 22us width, 20.00kHz. 20V DC input, 0.15A, 51V peak, 38V pulse, 20us width 25.25kHz. 22V DC input, 0.13A, 53V peak, 40V pulse, 20us width, 26.88kHz. 24V DC input, 0.12A, 55V peak, 42V pulse, 20us width, 28.74kHz. 26V DC input, 0.11A, 57V peak, 44V pulse, 20us width, 30.12kHz. 28V DC input, 0.10A, 58.5V peak, 46V pulse, 19us width, 31.65kHz. Second set of tests with 5307 module fitted. 8V DC input, 1.22A, 48V peak, 28V pulse, 57us width, 4.65kHz. 10V DC input, 0.90A, 47V peak, 30V pulse, 50us width, 6.41kHz. 12V DC input, 0.72A, 48V peak, 32V pulse, 46us width, 8.13kHz. 15V DC input, 0.55A, 49.5V peak, 35V pulse, 42us width 10.31kHz. 20V DC input, 0.40A, 52V peak, 38V pulse, 38us width 13.7kHz. 22V DC input, 0.36A, 54.5V peak, 42V pulse, 36us width, 14.79kHz. 24V DC input, 0.32A, 55.2V peak, 43V pulse, 36us width, 15.92kHz. 28V DC input, 0.27A, 59.2V peak, 47V pulse, 34us width, 17.48kHz. Added extra 680uF across C6, module still fitted. 6V DC input, 1.37A, 47.2V peak, 26V pulse, 60us width, 3.5kHz. Removed extra capacitor and found no chirping. David |
14th Nov 2022, 8:33 pm | #66 |
Heptode
Join Date: Jan 2011
Location: Croydon, London, UK.
Posts: 772
|
Re: HP 5300B Overcurrent
That's a very useful set of result, David, Thank you.
On the faulty unit, it would be good to see what measurements you get when David's tests are reproduced, applying the DC supply to A1C6. Do the tests with the module removed, as you don't have a 5307 module. Stop when the peak voltage gets close to 65V, you don't want to burn out another 68V Zener. That gives 5 measurements at each supply voltage. Try using normal mode on your oscilloscope, rather than auto trigger and trigger on the positive slope, AC coupled. You ought to be able to get a steady trace if you adjust the trigger level control. Adjust the time per division to show no more than 2 or 3 complete cycles to make time measurements easier. You may not be able to measure the frequency directly using your oscilloscope. No matter, report the width of the positive pulse and the time between the peaks. The frequency can be calculated from that easily enough. Paula |
15th Nov 2022, 12:49 am | #67 |
Dekatron
Join Date: Sep 2005
Location: Seaford, East Sussex, UK.
Posts: 5,997
|
Re: HP 5300B Overcurrent
Is the A1T2 transformer still on flying leads? If so I suggest it is reinstalled properly and while you are there inspect the boards for dry joints.
When using the scope attach the earth as close to the Zener as possible. |
18th Nov 2022, 12:16 am | #68 |
Nonode
Join Date: May 2016
Location: Cornwall, UK.
Posts: 2,333
|
Re: HP 5300B Overcurrent
I made the tests this evening, gradually winding up the voltage across C6 and making notes. Things were going well, with the flyback peak as measured on the oscilloscope reaching a maximum of 58V and holding, even with 28V applied. The waveform then is shown attached, at 1V/div (x10) and 5µs/div.
Everything looked hunky dory, so I measured the voltage rails. +5.00V. Great! thinks I, with the others OK, until I reached the -5V rail which was sitting at +4V. I then touched T2 (which is reattached to the board), it was hot, there were sparks from beneath it and the oscilloscope probe melted to Q17. Subsequent tests with T2 removed suggest the previously possibly troublesome pin 4 is open circuit from the rest of the secondaries, though peculiarly all the other pins check out. The problem appears isolated to the pin 4 tap. Without cutting away the encapsulation and reconnecting it (if indeed that's all that's wrong) I can't see any way of keeping this from the 'small electricals'. This is very frustrating after all the kind responses and thoughtful analyses, and factory going well beyond the call of duty in taking the careful time to make all the measurements. Also since I went to the trouble of testing T2 with advice from Ed Dinning, and it was fine, thinking now it was perhaps part of the problem all along is most disappointing. At least the original +5V rail issue appeared to have been solved. |
18th Nov 2022, 12:25 pm | #69 |
Dekatron
Join Date: Sep 2005
Location: Seaford, East Sussex, UK.
Posts: 5,997
|
Re: HP 5300B Overcurrent
It's running at 40KHz but the trace looks much better and there is no indication of a fault condition which would result in distortion at the top of the square wave. The -5V tap is taken from the -17V winding, does that show continuity? T2 looks like a ferrox pot core that has been encapsulated.
Rather than scrap it, I would be interested in the challenge. |
18th Nov 2022, 6:45 pm | #70 |
Heptode
Join Date: Jan 2011
Location: Croydon, London, UK.
Posts: 772
|
Re: HP 5300B Overcurrent
After a long battle it seems that you have identified the remaining fault. As PJL says, it sounds like something that can be repaired. I suspect the termination was damaged when T2 was removed in the first place. It's certainly worth trying to repair it, or letting PJL have a go if you don't think you can do it yourself.
If nothing else, there have been a few lessons to take away from this exercise. The biggest is that you have to watch lead lengths in switch mode power supplies! Paula |
20th Nov 2022, 9:08 pm | #71 |
Nonode
Join Date: May 2016
Location: Cornwall, UK.
Posts: 2,333
|
Re: HP 5300B Overcurrent
Thank you for the further encouragement! Faced with a multi-pin encapsulated transformer with an obsolete HP part number, I thought that was curtains.
I appreciate the sentiment, PJL. This unit has already surpassed my knowledge many times so the challenge barrier has receded into the mists long ago! With a house to build, a 6" desk when the keyboard's pushed away and two sets of grandparents' possessions piling up in a 3.5m wide building site, you can understand my wariness for rewinding what looks to me a jolly complicated little transformer. frsimen - what's so confusing is that T2 checked out. I even Meggered the contacts and the insulation was perfect. What's the plan then? Saw off the encapsulation? I think I'm going to need even more hand-holding for this bit... |
21st Nov 2022, 6:26 pm | #72 |
Heptode
Join Date: Jan 2011
Location: Croydon, London, UK.
Posts: 772
|
Re: HP 5300B Overcurrent
I've never seen one of these in the flesh, so I don't know how difficult it will be to repair the transformer. It sounds like there is a poor connection to pin 4, although the sparks that were seen under the transformer are a rather concerning. A little careful removal of the encapsulation near pin 4 may gain you access to the top of the connection, so that it can be checked. Sawing off the casing sounds a bit drastic, I suggest.
Be very careful how you remove the transformer, as you don't want to damage the connections to the other pins. Maybe PJL or others can offer some better advice. Paula |
24th Nov 2022, 7:22 pm | #73 |
Nonode
Join Date: May 2016
Location: Cornwall, UK.
Posts: 2,333
|
Re: HP 5300B Overcurrent
This is the object in question. Encased in hard plastic, possibly potted with some compound from the bottom.
The board contacts are now eyelets, soldered to the tracks. |
25th Nov 2022, 8:50 pm | #74 |
Heptode
Join Date: Jan 2011
Location: Croydon, London, UK.
Posts: 772
|
Re: HP 5300B Overcurrent
About 15 years ago I had a problem with a small pulse transformer in an HP spectrum analyser. From memory, it was of a similar construction to this transformer. I had to break the outer case to gain access to the tiny bobbin within and that was damaged during my efforts. Fortunately, a new pulse transformer wound on a ferrite bead restored normal operation that time.
Based on my previous experience, I can only suggest that you are careful with whichever method you use. If it were mine, I think I would try to separate the potting from the outer case, working from the bottom, in the hope that the outer case could then be removed intact. Others might have some better suggestions. Paula |
26th Nov 2022, 9:50 am | #75 |
Octode
Join Date: Jul 2009
Location: Cheshire, UK.
Posts: 1,733
|
Re: HP 5300B Overcurrent
No idea if it would work in this case, but I used to repair potted transformers at work, the potting compound could be softened with a heatgun, then carefully scraped out around the connections, to access the hidden solder joint, which was dry jointed. To refill araldite was used, also glued the side back on, they were round transformers potted into a cube.
David |
27th Nov 2022, 10:51 am | #76 |
Dekatron
Join Date: Sep 2005
Location: Seaford, East Sussex, UK.
Posts: 5,997
|
Re: HP 5300B Overcurrent
The approach to this will depend on the fault. You need to check the continuity of all of the windings to confirm it is only pin 4 that is disconnected. Also, carefully check the PCB for damage.
Is pin 4 loose? Acetone might soften the filler. |
27th Nov 2022, 1:35 pm | #77 |
Octode
Join Date: May 2009
Location: Linkoping, Sweden
Posts: 1,465
|
Re: HP 5300B Overcurrent
If I read the thread properly you said that you had contact between pins 3 and 5 supplying the -17V connection, but not to pin 4 from any other pin.
If so you might be able to repair the number 4 connection by applying heat with your soldering iron to the pin and gently pressing on the pin with the solder tip - don't do it for too long and try a few times checking the connection in between. This might melt the solder that the transformer wire has been soldered with to the pin enough to make proper contact again. Another solution, not original though, which might work would be to use a -5V negative voltage regulator like the LM7905 hooked up to the -17V connection and then to the -5V connection. Input to the LM7905 should go to the junction of L5/C21 and output to L4/C25 and also GND to COMMON, and you should at least lift one side of CR4 if the transformer connection to pin 4 becomes intermittent which will not be OK for the LM7905. This will only work if the -17V connection can supply enough current for both the -5V and -17V rails.
__________________
Martin, Sweden |
8th Dec 2022, 9:22 pm | #78 |
Nonode
Join Date: May 2016
Location: Cornwall, UK.
Posts: 2,333
|
Re: HP 5300B Overcurrent
Thanks all. I've dug away at the encapsulation and found that there was some kind of sleeve on the pin which had intermittent contact with pin 4 unless the pin was pressed in a certain way. I've soldered them together and have reliable contact.
However, A1Q17 is deceased and I don't know what to use for a replacement. I've looked through all the HP Bench Briefs cross-check lists I can find and there's no result for 1854-0487, described in the manual as 'transistor: Si NPN'. I have several of the right form factor to hand that my TITS says are NPN. Would a BDW93C be adjudged a drop-in replacement, for instance? |
9th Dec 2022, 12:24 am | #79 |
Heptode
Join Date: Jan 2011
Location: Croydon, London, UK.
Posts: 772
|
Re: HP 5300B Overcurrent
Well done on the transformer repair. I don't think the BDW93C is a suitable replacement for Q17 as it is a Darlington transistor.
Possible substitutes are the TIP41C, BD241C or BD243C. All have a Vce of 100V and maximum collector currents of 5-6 amps. Of these, the TIP41C seems to be readily available from reputable suppliers. Paula |
11th Dec 2022, 12:22 am | #80 |
Nonode
Join Date: May 2016
Location: Cornwall, UK.
Posts: 2,333
|
Re: HP 5300B Overcurrent
Ah, OK. I misread the book! I have used a BD241C as I had one from a Uher power supply detailed elsewhere on the forum. Alas, the story's the same. With DC across C6, up to 26V, there's a good switching waveform on the BD241C collector. In fact, it doesn't peak as high as before so is further away from the 68V zener breakdown. Voltage rails were OK, although with -5V still ~0.5V low. However, applying mains burns out the zener once again.
|