SWR California Blonde '98

[lowell]

I believe this is from 1998. It came in with transistors different from those listed on the schematic. Not on any of the schems I've seen. It had 2SC2837 outputs. Anywho, one had shorted. I replaced both outputs with a bit beefier transistor - 2SC5242. NPN and same pin out, slightly higher rating.

Now, I'm mounting them to the heatsink, but there are no insulators! I checked the old transistors and the heatsink etc... there are NO insulators in here. This heatsink is mounted directly to the chassis. Looking at the schem this seems very odd and crazy to me that this amp was working at all previously. It WAS working when it came in but customer said the power would go in and out. I didn't believe him as I ran the thing for hours on 11 and had no issues. Finally I jiggled the ext. spkr jack while it was running and it blew the fuse, and shorted one of the outputs. So I assumed one of the outputs was the issue and had been on its last leg and causing the issues for the customer.

Now I'm wondering how it was working at all with the collectors of the outputs shorted to ground via the heatsink. This DOES NEED INSULATORS correct? Just want to be sure - I can't see how I could be wrong - but safer to ask you all.

thanks

http://music-electronics-forum.com/a...-blonde-00.pdf

[lowell]

Nevermind... I began cleaning the old thermal paste off the heatsink and voila...insulators. Duhhh...

Well in any case I replaced both drivers, the Vbe multiplier, and both outputs.

With my bulb limiter:

It seems my negative rail is good, but somehow there is only -9.5v reaching the 180ohm neg rail dropper. There IS -32v at the low v supply 5W dropping resistor however. Odd. The pos rail is +32v.

Guess I'll need to investigate the underside of the board as the schematic doesn't seem to be exact. That neg rail low v 5W resistor is not showing continuity to the 180ohm high v dropper. If I short them with a wire, the neg rail is still low around -25v.

[lowell]

Simulpost!

[g1]

Was the output that blew on the neg side? Could it have damaged the trace?
And on a separate issue, I'm not clear on what happened with the ext. speaker jack but it sounds scary. Could it have a fault that is shorting the output?

[lowell]

I can't remember which side. It's working now. There was a solder short on one of the drivers I installed...oops.

I am not sure what happened honestly...gonna test it again now that it's working.

NOTE* THE 180ohm and 270ohm on the negative rail in the schematic are swapped. This looks factory soldered. So either the schem is wrong or they installed em wrong. In any case its working fine and bias is symmetrical.

So the 2SC5242 datasheet says typical Vbe is 1v. Is that what I set the bias pot to? Or do I want some voltage reading across the emitters thus giving a specific quiescent current?

[lowell]

I just revisited Teemuk's awesome book. http://www.thatraymond.com/downloads...ttala_v1.0.pdf

So I realized I've come quite a long way over the years when reading this now. I can mostly understand what he is saying haha... so beta and hfe are synonymous correct?

I replaced the bias servo (Vbe multiplier etc..) 2SC4153 https://www.semicon.sanken-ele.co.jp...4153_ds_en.pdf in this amp with a MJE15028G http://www.mouser.com/ds/2/308/MJE15028-D-110804.pdf. Now, reading the book I think this is a mistake. The 2SC4153 has a hfe or 70-220. The MJE15028G has hfe of 2/3. Wow, big difference. So what I'm understanding here is that the 4153 is far more sensitive to temperature and will be far better at handling thermal runaway. Is this correct?

Edit. Just set them at 10ma. (1mv across .1Re) The Vbe is at 540mv. And the amp sounds good...clean.

[lowell]

Also, in regards to thermal pads. I bought a sheet of stuff on mouser a while back. It's like foam rubber. About 1/32 to 1/16" thick, grey material that is double side sticky. Is it safe to use this to mount TO247 to? I'm worried it's not good for high power applications and that maybe mica is the correct choice.

And again, where do you get TO-247 mica pads from?? I can't find em on mouser or digi.

[Enzo]

Sil-pads are the silicone sheet stuff, at Mouser enter sil-pad in the search. When the sort items come up, the material is eithe Kapton or SIlicone. Sil-pad is a brand name for Bergquist products, I don;t know the other brands. Pick the size you want.

Digikey, search Thermal-pads sheets. various sizes. They have a large selection. I imagine they make adhesive sil pads, but I've never used them.

I would describe the sil pads as rubbery, but not spongy, but that is just words. They are the typically gray things under transistors, though some are beige or tan or pink or green. I'd be very surprised if you have never seen them up close.

[g1]

And again, where do you get TO-247 mica pads from?? I can't find em on mouser or digi.

Interesting, lots for TO-220 and TO-3, but hard to find mica for TO-218/T0-247.
Here's some:
TO-247MICA - NTE ELECTRONICS - MICA FOR TO-247 CASE | Newark element14

[Leo_Gnardo]

Interesting, lots for TO-220 and TO-3, but hard to find mica for TO-218/T0-247.
Here's some:
TO-247MICA - NTE ELECTRONICS - MICA FOR TO-247 CASE | Newark element14

Good find, that. I've sometimes taken scissors to TO3 mica wafers, these are much better. Will have to invest in a 100-pak.

[lowell]

Thanks. Enzo I'm def familiar with sil-pads. However this stuff is spongy. It's not a thin rubber "pad". Its far thicker than a mica insulator or sil-pad. I'll see if I can track down what it is in my previous orders.

Thanks for the links guys. Seems very odd to me that TO-247 are still being sold and don't seem to be going anywhere...but the insulators are so unavailable.

[lowell]

Also any comments on the Vbe multiplier in my previous post? I ised a MJE15028G to replace the C4153. Doubt I made the right choice there after reading Teemuks book.

Think I'm on the right path but some confirmation wont hurt. Seems Beta is a key spec for a temperature sensitive bias transistor.

[Enzo]

The insulator is electrical, not thermal. We want heat to transfer as freely as possible from the part to the heat sink. We don't want puffy, spsongey stuff in between. That would act as a thermal insulator, the opposite of what we want.

You are over analyzing the transistor. First, I have no idea what an hfe of 2/3 means. In your data sheet, I see a 2 and a 3 under hfe linearity, but look one line up, and see hfe for the part is a minimum of 40.

This is just a bias transistor, it establishes a voltage difference and holds it.

[Jazz P Bass]

The fact that that 'puffy' stuff is double sided sticky concerns me.

Good luck getting the transistors off if need be.

[g1]

Can you link to the spongy stuff you bought off Mouser ?