23-08-2017, 04:27 PM
Running up the LOP board on an external supply suggested a short circuit or something else badly wrong. So I disconnected the power feed to the output stage and things came to life. There's drive to the output transistors so I connected power to the output stage via a 100R resistor. Obviously taking too much current. One of the R1039 output transistors is duff, with a base-emitter short. What's a modern replacement? Need to replace both? All I have here are some BU508A which are in a tab package, BUX80 which may be OK at 400V rating and BU102 which I'm pretty sure aren't up to the job. I can't find any data on the R1039 but at a guess it's taking no more than 250V. Other parameters will be critical too. I suspect BU105 would be a good replacement but I don't have any.
Here's the reasoning. The supply to the LOP stage is nominally 53V. If 53V will do a forward stroke in approx 52us then for 12us approx flyback it will take 53V * 52/12 = 230V. I have a feeling I've missed a factor of 2 somewhere so that could be 460V overall. Split between 2 devices in series so around 230V per device. Not very demanding in modern terms but it's not just voltage rating. Current gain and switching time matter a lot. In theory I could dig out my Telequipment transistor curve tracer which hasn't been plugged in for over 20 years but I suspect that would give me more grief.
The other LOP card (the original one in the set which has been somewhat got at) has the early square LOP transistors, type D1418. These at least both show as OK on cold test with a DMM. So I think it's time for a close inspection of the board, fix any obvious damage and run it up.
Oh, and just for fun, the PM5519 has got a great looking hum problem. The circle has lumps out of it that are moving at the difference between it's field frequency and mains. That's for another day, it's working well enough for now.
Here's the reasoning. The supply to the LOP stage is nominally 53V. If 53V will do a forward stroke in approx 52us then for 12us approx flyback it will take 53V * 52/12 = 230V. I have a feeling I've missed a factor of 2 somewhere so that could be 460V overall. Split between 2 devices in series so around 230V per device. Not very demanding in modern terms but it's not just voltage rating. Current gain and switching time matter a lot. In theory I could dig out my Telequipment transistor curve tracer which hasn't been plugged in for over 20 years but I suspect that would give me more grief.
The other LOP card (the original one in the set which has been somewhat got at) has the early square LOP transistors, type D1418. These at least both show as OK on cold test with a DMM. So I think it's time for a close inspection of the board, fix any obvious damage and run it up.
Oh, and just for fun, the PM5519 has got a great looking hum problem. The circle has lumps out of it that are moving at the difference between it's field frequency and mains. That's for another day, it's working well enough for now.
www.borinsky.co.uk Jeffrey Borinsky www.becg.tv
