Like I said, very busy making my Zardoz costume, so forgot to clean inbox.
Thanks! Have a friend with a workshop, I’ll ask him for help.
Also, another friend has a 3D printer. Gonna ask him to make PL/ZM style brackets for FE. Expect fugly amp in a little while. Need to finish linestage first. Now to continue PSU plan with schematic. It will be a fugly schematic, on paper. I like.
Dennis: I will go for those Hammonds u recommended. Rest of PSU will be P2P, connected with metal plates, old style.
Time to put you 0F23s to use soon?
Thanks! Have a friend with a workshop, I’ll ask him for help.
Also, another friend has a 3D printer. Gonna ask him to make PL/ZM style brackets for FE. Expect fugly amp in a little while. Need to finish linestage first. Now to continue PSU plan with schematic. It will be a fugly schematic, on paper. I like.
Dennis: I will go for those Hammonds u recommended. Rest of PSU will be P2P, connected with metal plates, old style.
Time to put you 0F23s to use soon?
Hi, guys!
Finishing my Mouser basket, I am considering ordering some alternate values for R3-4 in order to increase bias currents of the FE JFETs from approx 5-5,5ma, to approx 8ma as per spec (Nelsons article). These are 10R per now. Maybe a stupid question, but do I increase or decrease the values iot achieve this? And what should I look out for? Any risks? 5-5,5 is in the low range, 8 is spec and safe as far as I can tell.
Thinking about 7-8R. Is that good or wrong thinking?
Any insight is very appreciated.
Cheers,
Andy
Finishing my Mouser basket, I am considering ordering some alternate values for R3-4 in order to increase bias currents of the FE JFETs from approx 5-5,5ma, to approx 8ma as per spec (Nelsons article). These are 10R per now. Maybe a stupid question, but do I increase or decrease the values iot achieve this? And what should I look out for? Any risks? 5-5,5 is in the low range, 8 is spec and safe as far as I can tell.
Thinking about 7-8R. Is that good or wrong thinking?
Any insight is very appreciated.
Cheers,
Andy
Found heatsinks for the JFETs! Anyone have any experience with these or the likes?
Goal: less temp sensitive JFETs. And hence maybe less offset drift at the FE. Achieveable?
https://no.mouser.com/ProductDetail/Aavid/575200B00000G?qs=VJtzl2NKxc3oHvqgDJX11Q==
Andy
Goal: less temp sensitive JFETs. And hence maybe less offset drift at the FE. Achieveable?
https://no.mouser.com/ProductDetail/Aavid/575200B00000G?qs=VJtzl2NKxc3oHvqgDJX11Q==
Andy
smaller source resistor - higher current
full Idss when resistor is 0
//////////
better cooling, lesser temp delta
full Idss when resistor is 0
//////////
better cooling, lesser temp delta
Haha, full IDSS and blown parts. I’ll check out 7-8R then. I will also order some alternate values for R10-11 and 12, iot experiment with higher bias on the outout end of the FE, cause I might get hold of some machined big alu sinks in time. Targeting 75-100ma.
ZM: what values to choose? I guess I’ll loose some PSRR reducing R10-11 values. So how high can I go with bias with original values without the gain getting too high or voltage losses too great? Would you increase over 45ma at all without using different values?
ZM: what values to choose? I guess I’ll loose some PSRR reducing R10-11 values. So how high can I go with bias with original values without the gain getting too high or voltage losses too great? Would you increase over 45ma at all without using different values?
I wouldn't go for more than 50% above prescribed mosfet Iq, if at all
for JFets, no need or reason to go above 7-8mA
for JFets, no need or reason to go above 7-8mA
😀 Cool. I’ll keep to specs as far as possible and add a couple of alternative R13s to adjust gain wrt future speakers. Also I added a set of R3-4s 7R5 to test 7-8ma bias on the JFETs.
And look! I found these: https://no.mouser.com/ProductDetail/Aavid/575200B00000G?qs=VJtzl2NKxc3oHvqgDJX11Q==
😀
And look! I found these: https://no.mouser.com/ProductDetail/Aavid/575200B00000G?qs=VJtzl2NKxc3oHvqgDJX11Q==
😀
Last edited:
2Sj313/2SK2013 Insulator ? Do I need to insulate these from heatsinks as I get no conductivity from case to sink?
I did it for mine just to be extra careful, but you don't need to do it.
Edited to add - LOL! What 6L6 said. 😀
Edited to add - LOL! What 6L6 said. 😀
I did it too. Fiddle so much around in there and did have that small fire, reduce risk where I can however miniscule 😀
Hi guys!
So two of these are now in production at Toroidy, headed my way in a little while:
TTSA0300 - Transformer AUDIO TSA300VA - voltage to 50V - Shop Toroidy.pl
And then I removed the last bit of hum (more or less). It just took a bit of reduction in bias to more normal and safe levels. That, and following Bonsais advice, got me there in the end.
But well well, since they are on the way, I might as well start putting my new PSU plans into life: dual mono porn cap old-style/Choky-style PSU with big aluminium plates for the ground plane, and inductors... wish me luck! Double todays capacitance, inductors and shared load between two psu’s, and those transformers, make ripple a factor only relevant to my scope. I hope. 😀
Just waiting for Choky to approve the PSUD model. Hoping to get to it before come summer. We’ll see, playing nicely so no rush.
And Russel: I have done some thinking, usual OCD me. I guess differences between yours and mine build wrt hum and no hum, come down to some relevant factors:
1: Your tranny is bigger, badder and likely better than mine
2: you have 90kUf more capacitance for less load (you run lower bias than me).
But if yours is silent to 100db, well, I’ll be inspired. My first goal is 97db.
Cheers from corona-closed Norway!
Andy
So two of these are now in production at Toroidy, headed my way in a little while:
TTSA0300 - Transformer AUDIO TSA300VA - voltage to 50V - Shop Toroidy.pl
And then I removed the last bit of hum (more or less). It just took a bit of reduction in bias to more normal and safe levels. That, and following Bonsais advice, got me there in the end.
But well well, since they are on the way, I might as well start putting my new PSU plans into life: dual mono porn cap old-style/Choky-style PSU with big aluminium plates for the ground plane, and inductors... wish me luck! Double todays capacitance, inductors and shared load between two psu’s, and those transformers, make ripple a factor only relevant to my scope. I hope. 😀
Just waiting for Choky to approve the PSUD model. Hoping to get to it before come summer. We’ll see, playing nicely so no rush.
And Russel: I have done some thinking, usual OCD me. I guess differences between yours and mine build wrt hum and no hum, come down to some relevant factors:
1: Your tranny is bigger, badder and likely better than mine
2: you have 90kUf more capacitance for less load (you run lower bias than me).
But if yours is silent to 100db, well, I’ll be inspired. My first goal is 97db.
Cheers from corona-closed Norway!
Andy
Last edited:
hi Andy,
what is the expected delivery time from Toroidy?
In January was 4-5 weeks...
I'm thinking to buy from Toroidy other stuff
what is the expected delivery time from Toroidy?
In January was 4-5 weeks...
I'm thinking to buy from Toroidy other stuff
Hi Andy and ZM -
ZM - Thank you! I was absolutely scratching my head when looking at the article on the FW site vs. your referenced post. R8 and R9 in the article actually refer to R10 and R11. Making much more sense. I am still weakling re: actual circuit, but each day learning a bit more.
The FE bias doesn't drift too badly, I may just be worrying about nothing. It does stabilize after an hour or so. It is just challenging to get set within 5% on each side. It wanders about 5mA against a 45mA target when the lid comes off. The DC offset drifts 20-30mV immediately when the lid is off. Targeting 0V. I have it centered at about 8mV on one side and 6mV on the other.
First action will be larger heatsinks for Q3 and Q4 just to sleep better. They are really toasty, but not out of SOA. I think (as usual) that I am worried about nothing, but larger sinks will give those devices some help either way.
Andy - your observations and bias / offset plan are very similar to mine. I let it "cook" for about 2 hours initially or just wait for the meters to stabilize for small tweaks with the lid on. I use 6 meters for the front end (2x R10 2x R11 2x between "Top of R12" at C3 junction and GND). See pic. I use 4 for the output stage (2x at source resistor and 2x at output/GND)
I suppose I need 10 meters 😀
The output stage is absolutely rock stable. I can dial the bias to within a mA and null the offset to within 0.1 mV. Easy-Peasy.
Pics. Of course! Wanna see my kids? 😀 Here's BA-3 showing off. To the left is BJ2, and hiding shyly below BA-3 is Balanced Iron Pre.Current state of affairs after letting it sit over night at 0A4 bias per output device. Sinks are getting toasty, but not true burning amp toasty yet.
Thank you, gentlemen!
View attachment 930540
View attachment 930541
Question about the UMS and sinks: did you drill new holes yourself to fot the two BA-2 OS’es and fets, or was it more or less a direct fit?
PS: if you care, see my thread «BA-3 revisited». I am also making dual mono, and am considering letting myself inspire from your 6 deep config. Nice work!
Regards,
Andy
Hi Andy -
That's very kind. Thank you. It was one of my most ambitious projects to-date, and it was a tough decision for me to modify an existing working amplifier. I think I've finally reached "Burning Amp" with it. Can't wait to try out balanced mono-blocks later down the road along with the BA-1 output stage.
A portion of the Toshiba devices I got to modify my input stage were likely fakes 🙁, so I will contact your recommended seller for some known good pairs. A bit of a delay, but no sweat. The current Fairchild parts are just fine, but I wanted the "upgrade" to be the best.
For the 6-Deep configuration, I used the "Toxic Masculinity" chassis. Thread linked below. They are discussing another group buy.
Pico's Toxic Masculinity Class A Amp Chassis
It is UMS, 4U/500, and has enough holes tapped for 12 devices per side. This application was part of the discussion around 500mm deep. So, it is a drop-in perfect fit. I have not mastered drilling and tapping with precision as yet. So, I wanted heatsinks that were properly done first.
There are also a few (what I'd consider) easy ways to get the 12 deep into a 400mm chassis if you're using 5U/400. Itsmee posted a wonderful idea that I would have used had the 500mm not become available.
Not sure what chassis you chose or are thinking about. However, keep in mind that I'm running ~24V rails. If you're going to run 32V rails (I think I read that), then you will likely need substantially more heatsinking that I do. 5U/500 would not be out of the question.
I did a few calculations first (with some help from ZM and Dennis if I recall) to see where I wanted to set Iq to match voltage and current into a load along with keeping the heat dissipation as low as possible while still achieving a power output target. Since it's push-pull, it wasn't quite as challenging to get where I wanted, but it was an excellent learning exercise.
Following your thread with interest. Have fun!!!!
That's very kind. Thank you. It was one of my most ambitious projects to-date, and it was a tough decision for me to modify an existing working amplifier. I think I've finally reached "Burning Amp" with it. Can't wait to try out balanced mono-blocks later down the road along with the BA-1 output stage.
A portion of the Toshiba devices I got to modify my input stage were likely fakes 🙁, so I will contact your recommended seller for some known good pairs. A bit of a delay, but no sweat. The current Fairchild parts are just fine, but I wanted the "upgrade" to be the best.
For the 6-Deep configuration, I used the "Toxic Masculinity" chassis. Thread linked below. They are discussing another group buy.
Pico's Toxic Masculinity Class A Amp Chassis
It is UMS, 4U/500, and has enough holes tapped for 12 devices per side. This application was part of the discussion around 500mm deep. So, it is a drop-in perfect fit. I have not mastered drilling and tapping with precision as yet. So, I wanted heatsinks that were properly done first.
There are also a few (what I'd consider) easy ways to get the 12 deep into a 400mm chassis if you're using 5U/400. Itsmee posted a wonderful idea that I would have used had the 500mm not become available.
Not sure what chassis you chose or are thinking about. However, keep in mind that I'm running ~24V rails. If you're going to run 32V rails (I think I read that), then you will likely need substantially more heatsinking that I do. 5U/500 would not be out of the question.
I did a few calculations first (with some help from ZM and Dennis if I recall) to see where I wanted to set Iq to match voltage and current into a load along with keeping the heat dissipation as low as possible while still achieving a power output target. Since it's push-pull, it wasn't quite as challenging to get where I wanted, but it was an excellent learning exercise.
Following your thread with interest. Have fun!!!!