Any chance you can post some picks of the iron? I'd be curious. It's also possible that if identified you can optimize the circuit for that particular transformer.
Gillespie triode Williamson build
- Thread starter derekva
- Start date
Ah. Sorry - I misunderstood the question.
Can you send a couple of photos and part numbers (should be stamped on the bell housings)?
Its Chicago BO-13, thats all the numbers I have. Catalog specs as follows:
primary 10K ohms
screen grid taps for extended power (no clue what %)
secondary 4/8/16 ohms
20 watts
807, 5881 tubes
Thats straight out of their 1958 transformer catalog. The other one I have looks a little nicer, but you can get a feeling for how fugly the amp is. Its actually nicely built, but someone was not a painter.
primary 10K ohms
screen grid taps for extended power (no clue what %)
secondary 4/8/16 ohms
20 watts
807, 5881 tubes
Thats straight out of their 1958 transformer catalog. The other one I have looks a little nicer, but you can get a feeling for how fugly the amp is. Its actually nicely built, but someone was not a painter.
crispycircuit
Super Member
If you can find a Sam's schematic. They often list replacement transformers for that unit.
They will in fact work well in a Mullard design. It's interesting however: phase shift issues in true Williamson designs (using all four stages) occur primarily at the low end of the spectrum, where the phase shift of the coupling caps from the extra stage of amplification come into play. In true Mullard designs (where a pentode drives the CC inverter), it's the high end where phase shift issues primarily occur, as the very high output impedance of the pentode stage allows Miller to have full effect in the inverter stage -- particularly when a 12AX7 is used for the inverter as is the case in the original Mullard design. As a result, the open loop high frequency response of the original Williamson design is actually quite decent with all its med Mu triodes and low value plate load resistors, whereas the open loop high frequency response of the original Mullard design is by comparison quite poor. It can certainly be helped by using a lower Mu inverter tube and a triode AF Amplifier stage, but in any case, after proper NFB and response tailoring networks are applied, the overall response of the two different topologies (both at the low and high frequency end of the spectrum) can be made to be nearly identical.
Dave
Dave
The 5881 version is also based on the W4AM iron. I applied it and it worked out very well with a slight tweak to one cap value to cure a slight rising response on the top end. Without the two transformers side by each to directly compare I'll never know for certain that it is the same, but I suspect the item in the can has a lot in common.
Thnx for response!
One more, about the Marantz Mullard implementations?
Last edited:
indosmokejon
Active Member
PM sent.
I just finished a pair of these. They're very nice. I found a lone 51-17 OPT on eBay for $75 and a "fixer-upper" W4 for $150 that had a good OPT. I used two Stancor power trannies I bought on eBay for $120 for the pair and some Hammond 8H/150ma chokes from Hawkeye Electronics for under $100 (Hawkeye has *great* prices on Hammond iron, BTW). Coupling caps are Cornell-Dubilier 1000v 942C series, which are a tremendous bargain, IMO. The two OPTs look identical but are stamped differently. I didn't worry about it too much. I tried the parallel 220K resistor across one driver plate resistor but didn't care for how it sounded so I left it out.
The power trannies are 400-0-400 because I prefer a 5U4 rectifier. Voltages dial in perfectly. The sound is clean and lively. Since 12wpc is plenty for my old ProAc Responses, I could live with these for a long time.
The power trannies are 400-0-400 because I prefer a 5U4 rectifier. Voltages dial in perfectly. The sound is clean and lively. Since 12wpc is plenty for my old ProAc Responses, I could live with these for a long time.
The notes are just what's visible on the schematic - there is no reverse side (it came in .JPG form). If Dave has additional notes on the back side of the schematic that he wishes to photo and post, that'd be fine, but unfortunately, I don't have that image.
Sorry!
indosmokejon
Active Member
Derekva. Any chance you would share a BOM if you have one? I still don’t even have any amps to convert but I’m def gonna get some. I’m gonna put in an order soon anyway and I thought I would scoop this stuff up too.
Sure. I'll have to dig up my old Mouser order.
Ah. Good point.
One minor note - the schematic cut off the feedback loop capacitor value - should be 24pF for Chicago or 15pF for Stancor OPT.
[EDIT 02.05.2020: I had the values backwards originally - thanks to Jon for pointing this out]
Last edited:
How do you tell the difference between the two? As I mentioned above, mine look identical except they are stamped differently. One has the model (51-17) and a part number (138839) etched horizontally along the bottom edge, and the other has a completely faded paper label and just "51-17" stamped vertically in the middle of one side. Any idea which is which--or are they both the same? Any clues would be welcome. ;-)
If I remember correctly (and realize this is just an educated extrapolation based on a very small sample size), the Chicago OPTs stamped the part number on the side while the Stancor had it on the top. I'm not familiar with the vertical stamp + label transformer, but it may have been either a very early Chicago or a later replacement transformer (probably Stancor). I realize that this is not much help.
@dcgillespie - any ideas?
