diytube.com

corndog71 wrote: I still have all of my notes so I can shoot you a copy of my drawing of the driver board if you're interested.

Yes please.

After going over my drawings and finding a few mistakes I want to fix them before reposting them. There's one other thing that's been bugging me. You can tell I kind of winged it as I built this because I used the power supply from the SCA35 instead of the ST35. I figured it didn't matter that much and as I said it sounds fantastic. Lately I've been wondering about pulling one of those caps out and adjusting the resistors to better match the ST35 circuit. I doubt it would affect the sound too much but it would give me a little piece of mind. It'll also give me a chance to take some better pics of my work. Stay tuned!

Here is the partial SCA-35 power supply I had been using with bypass caps.
http://i576.photobucket.com/albums/ss20 ... 56cee9.jpg

I took out one of the 45uF caps as well as the bypass caps and put the proper resistor in there.
http://i576.photobucket.com/albums/ss20 ... dda685.jpg

I also secured the ground wire a bit better and removed the unnecessary insulation. Yes, my board is turning into swiss cheese. As soon as I can find a replacement I will do so.
http://i576.photobucket.com/albums/ss20 ... 1f574a.jpg

Remounted!
http://i576.photobucket.com/albums/ss20 ... ca8361.jpg
Hmm... I should probably cover up those terminals by the bias pot. If I were to do it again I would flip that EFB board around so that the bias control is away from the high voltage circuit.
http://i576.photobucket.com/albums/ss20 ... 68055d.jpg

Here's the EFB circuit. I used this to mount the regulator to the bottom of the box. http://www.radioshack.com/product/index.jsp?productId=2102859

http://i576.photobucket.com/albums/ss20 ... G_6904.jpg

Initailly, when building this amp I was very much aware of the high voltages present. I designed the first power supply with that in mind. When deciding to add the TC caps and redesigning it I seemed to have dropped the safety ball a bit. In order for us all to learn from these rookie mistakes I'm highlighting some needed changes.

For starters I think I'm going to remount the caps so that the resistors are on the bottom and the ground runs higher.

Another area that caught my eye this morning was the high voltage connection to my EFB board. As you can see I have the 360K resistor mounted going up from the pot. This exposes the hot side of that resistor right next to the bias pot.
http://i576.photobucket.com/albums/ss20 ... 1dd93a.jpg
An easy fix is to simply swing it down towards the lower part of the board.

http://i576.photobucket.com/albums/ss20 ... 4f8f56.jpg

Here is a drawing of my star ground. Hopefully it makes sense. It's basically a 3-point terminal strip bolted to the chassis. The center point is the ground to which both the power supply and driver boards connect. The power supply ground wire also carries the EFB ground, speaker terminal grounds, and Red/Yellow wire from the PA774.

This is also the voltage divider for the filament supply.

Seems a bit crowded here but I did it this way in order to ease diassembly/reassembly.

http://i576.photobucket.com/albums/ss20 ... 6e026d.jpg

Well, I learned another lesson yesterday. Sometimes you pick the part that works best even if it's cheap. I lost another one of those 50 Ohm 5W Mills resistors. It went from 50 to 2MOhms.

Ordering the good old ceramic cast resistors... Gotta remount everything anyway.

Hey try a choke of around 21 ohms or less with 300 or more mA and around 2H. Edcor has them, and some say the keeping the DCR down in the PS by keeping the resistance low (21 ohms or less) helps the sonics. Besides a good choke is an improvement, and you can compensate for the slight B+ increase , by adjusting the Bias to keep the dissipation down.

Great documentation - thanks for posting this!

So I'm in the process of rebuilding my ST35. Some of you may be disappointed that I ripped it completely apart. Fear not! From the ashes of the past a new amp will be reborn!

In the midst of said transformation another error on my part was discovered. When I drew out my old bread board layout I got one part in each channel wrong. What I drew as 27K should actually be 1.3K as per DIYtube's ST35 mod. For the record, I was using the correct part but when drawing out the layout I must've looked at the resistor to figure out the value and misread it. My part is actually closer to 1.27K but the writing on the Dale resister didn't include the 1 at the beginning and led me to believe it was 27K. That's what I get for not confirming the value with my meter.

So one more time here's the corrected "old" layout. I think you guys will like the new look. Stay tuned!

http://i576.photobucket.com/albums/ss20 ... c1685c.jpg

I like that drawing of the grounds on the graph paper. It looks like the drawings in the old Dynaco and Heath manuals, a lost art!

I read somewhere before about the power ratings on the Mills wirewounds being a bit optimistic . . . an inductive part will be great in that spot anyway.

The original ST35 (excluding cathode bias section) used 60-40-20uF in its power supply. The diytube uses 120-120-120, Audio Regenesis uses 47-120-120. Is there an optimal figure? Or does it depend on who you're talking to! I use 70-80-20 in my original version. Maybe the driver stage could be larger? I read on another forum - or maybe it was this one - where it's advisable to use as little capacitance as possible for a quicker signal response. But most people suggest to load 'er up. Too much storage capacity is never enough.

Also, I like the idea of using the Clarity Cap TC's in place of the can cap of the original - using a bigger chassis of course. I read somewhere that the value of the film cap (Clarity Cap in this case) doesn't need to be the same as the electrolytic. For instance, would a 80 uF film cap be equivalent to a 120 electrolytic? I'm wondering what the ratio is.

Thanks for any insight on this.

Stunch wrote:The original ST35 (excluding cathode bias section) used 60-40-20uF in its power supply. The diytube uses 120-120-120, Audio Regenesis uses 47-120-120. Is there an optimal figure? Or does it depend on who you're talking to! I use 70-80-20 in my original version. Maybe the driver stage could be larger? I read on another forum - or maybe it was this one - where it's advisable to use as little capacitance as possible for a quicker signal response. But most people suggest to load 'er up. Too much storage capacity is never enough.

Also, I like the idea of using the Clarity Cap TC's in place of the can cap of the original - using a bigger chassis of course. I read somewhere that the value of the film cap (Clarity Cap in this case) doesn't need to be the same as the electrolytic. For instance, would a 80 uF film cap be equivalent to a 120 electrolytic? I'm wondering what the ratio is.

Thanks for any insight on this.

It does seem a bit all over the place doesn't it? I've learned a couple rules to go by:

1. the first cap should be smaller in order to minimize the initial turn-on surge.
2. Ideally, the cap connected to the output transformers will be your biggest cap. There seems to be some debate over how much capacitance is necessary. Some say too much can mess with time constants (which I still haven't learned about) while others throw a bunch on there and call it a day.

I based my cap choices both on what others were doing and what I could afford in TC caps. I wanted to try the TC caps as they have an extremely low ESR compared to any electrolytic. Also, since electrolytics do degrade over time I figured that's another reason to go with a big film cap. Plus, I just wanted to be different and try something new.

I've never read anything about ratios between cap types.

jgf wrote:I like that drawing of the grounds on the graph paper. It looks like the drawings in the old Dynaco and Heath manuals, a lost art!

Agreed!

If CAD software hadn't made me so lazy.....