Bypass capacitor around VBE multiplier is placed to ensure stability. Most of the time you will see a capacitor with value from 100nF to 10uF used as a bypass around VBE multiplier. In case of AX11 amplifier, Mile (Apex) used 100nF.
oh my... I see my deadline running! I will check if it is easy to add! (the smalest value)
We didn't use bypass on earlier Vbe and it seemed to work fine. Does it just smooth out the Vbe response? On my Pass M2, there is a huge 3300uF cap on the Vbe so that the amp takes a minute before full bias is reached. It uses a clever opto-coupler and precision voltage sources as the bias regulator. No tempco. Works real well.
I want to try this on my next class AB amp.
I want to try this on my next class AB amp.
That LED is in a bad spot. It can only really be utilized if the outputs are mounted under the boards otherwise it is in the way of mounting T7 to the heatsink. I suppose an SMD LED could be used under the board. The bypass cap needs to attach to the collectors of T1 and T2.
Every project needs a deadline (FH11/FH9) The version AFTER the last version
Now with LED and capacitor
The potential of good ideas on this forum is great! Nevertheless, I will do a cut now or I might still even contribute from the seniors home! :
I will again provide the complete set of files for the FH11 V1.07, this time including the dot marks for the silk screen AND provisions for a temperature compensated VBE multiplier stage.
The layout which I used as standard is at least 3 times build and that means measured/checked.
Nevertheless I added a diode; LED2, *hot-wired* (in series to VBE multiplier transistor) and a capacitor (C25) to give the possibility to test/implement the compensation for this layout.
So you have 4 different possibilities (with FH11 and FH9): 1 side or 2 side PCB, with compensation or original as proved and tested.
In case you ordered double-side PCBs, simply fit with the components printed on the top (silk screen)/BoM besides C25 and you will get the proven and tested amp design.
In case you ordered double-side PCBs and you want to try the compensation with the LED do the same , but interrupt the track between the both pads of LED2 on the top. Then you can fit the LED and C25 and the changed resistors/trimmer.
In case you use a one-side PCB, fit all components according to the silk-screen/BoM besides C25and jumper the 2 pad of LED2. (Don't forget the jumpers for the 2 missing tracks on the top).
In case you use a one-side PCB and you want to try the compensation with the LED do the same, but leave the jumper and fit the LED2 and C25 and change resistors/trimmer. (Here again: Don't forget the jumpers for the 2 missing tracks on the top)
About reliability, xrk971, even if the scale do not fit completely: this layout is as error free as Windows 10
but I checked it several times, so I hope really heavy errors can't be found any longer. But I would ask you to check at least BoM and schematic, if all used parts are ok.
Now with LED and capacitor
The potential of good ideas on this forum is great! Nevertheless, I will do a cut now or I might still even contribute from the seniors home! :
I will again provide the complete set of files for the FH11 V1.07, this time including the dot marks for the silk screen AND provisions for a temperature compensated VBE multiplier stage.
The layout which I used as standard is at least 3 times build and that means measured/checked.
Nevertheless I added a diode; LED2, *hot-wired* (in series to VBE multiplier transistor) and a capacitor (C25) to give the possibility to test/implement the compensation for this layout.
So you have 4 different possibilities (with FH11 and FH9): 1 side or 2 side PCB, with compensation or original as proved and tested.
In case you ordered double-side PCBs, simply fit with the components printed on the top (silk screen)/BoM besides C25 and you will get the proven and tested amp design.
In case you ordered double-side PCBs and you want to try the compensation with the LED do the same , but interrupt the track between the both pads of LED2 on the top. Then you can fit the LED and C25 and the changed resistors/trimmer.
In case you use a one-side PCB, fit all components according to the silk-screen/BoM besides C25and jumper the 2 pad of LED2. (Don't forget the jumpers for the 2 missing tracks on the top).
In case you use a one-side PCB and you want to try the compensation with the LED do the same, but leave the jumper and fit the LED2 and C25 and change resistors/trimmer. (Here again: Don't forget the jumpers for the 2 missing tracks on the top)
About reliability, xrk971, even if the scale do not fit completely: this layout is as error free as Windows 10
Attachments
Abletec interest
I see some interest here in the Abletec +/- 53V power supply, model ALP0400-5301-D
I just stumbled across KestPart's listing for $19.95, which I don't recall being mentioned. Shipping to my home in North Carolina puts it up over eBay's $31.35 (shipped, which I purchased earlier today). However, if your purchase is over $50, the shipping is free... Kinda like "get three for the price of two"... Just a thought.
Disclaimer: I haven't dealt with KestParts, and cannot vouch for them.
I see some interest here in the Abletec +/- 53V power supply, model ALP0400-5301-D
I just stumbled across KestPart's listing for $19.95, which I don't recall being mentioned. Shipping to my home in North Carolina puts it up over eBay's $31.35 (shipped, which I purchased earlier today). However, if your purchase is over $50, the shipping is free... Kinda like "get three for the price of two"... Just a thought.
Disclaimer: I haven't dealt with KestParts, and cannot vouch for them.
Last edited:
I originally got my Abletecs from Parts Express with free shipping for $20 if part of $100 total order. I got two and some speaker drivers. It was a great deal. You need to read the thread for tips on how to use it. It needs to have a small load on auxiliary voltage supplies to work properly.
Eaglemaster,
Looks very nice. But if using LED in Vbe the resistors need to be changed following diagram provided by Still4given.
Looks very nice. But if using LED in Vbe the resistors need to be changed following diagram provided by Still4given.
Ty you for advice, xrk971, but design is made for your original FH11, only with the OPTION to implement the temperature compensation! As long as it is not proved, that the FH11 with temperature compensation is really working better than the original one I would not like to change silk screen or BoM. And if you had a look on it (schematic) you can see that the LED right now is still hot-wired!
As you told, you think about ordering some PCBs, then you can test if it works and if the values of still4given fit well in the FH11. As I could see you have the equipment to test it easily!
If the improvement and values are really sure, I would of course adapt schematic, BoM and layout.
Length of this thread!
I just had a discussion with a moderator about searching threads!
Beside some insufficiencies of the search engine one of the main problems is length of this thread.
I would suggest either to stop here completely and open a ' 100W Ultimate Fidelity Amplifier 2' or continue here only with already started themes and put new themes to the new thread.
Best would be if the owner of this thread would start the new one too!
I just had a discussion with a moderator about searching threads!
Beside some insufficiencies of the search engine one of the main problems is length of this thread.
I would suggest either to stop here completely and open a ' 100W Ultimate Fidelity Amplifier 2' or continue here only with already started themes and put new themes to the new thread.
Best would be if the owner of this thread would start the new one too!
I like it just as it is. By far my favorite thread. You can unsubscribe if you get too many notifications. Xrk has developed another thread to help folks find some of the Apex projects.
Don't get me wrong!
It is not about notifications, still4given. And it is not about my sensivities.
It is simply about the applicability of the search function.
Maybe you remember that Xrk had to open a second thread because first one got too long, though it is much shorter than this one!
But besides this: it was only a suggestion for better handling and I see your post as a first opinion to it.
It is not about notifications, still4given. And it is not about my sensivities.
It is simply about the applicability of the search function.
Maybe you remember that Xrk had to open a second thread because first one got too long, though it is much shorter than this one!
But besides this: it was only a suggestion for better handling and I see your post as a first opinion to it.
Yes, the search feature is too limited. What is your exact difficulties? Here's a few tips:
(1) Use the "Show attachment window" to search for attachments.
(2) Use CTR+F to search for a text in a page. Use F3 for search next.
(3) Display 100 posts per pages. So if you want to find post #17733 it will be on page #178.
(4) To go to page #178, edit the link on the address bar to e.g. "http://www.diyaudio.com/forums/solid-state/164093-100w-ultimate-fidelity-amplifier-178.html"
Many thanks Jay!
You just pointed out some of the insufficiencies of the search function!
And you provided 'workarounds'. Thank you again for that!
Searching attachments and getting to searched posts had been 2 points!
Another point had been that there is no 'help-function' for search with all those points you mentioned here anywhere in a thread! (The system should provide it on the searchpage)
Nevertheless, even if no other limitations: too long threads slow down search funktion!
And due to the many variants or alternatives of some designs a good search-function is essential.
Last edited:
Perhaps a thread like this one is not for everyone. For a builder, tester like myself, it is a candy store. Always a new flavor popping up to tickle my taste buds. If you start opening a new thread for each new design, the threads quickly fall out of sight and it becomes very difficult to remember where to find the information you are looking for. Mile has three basic threads. This one, his Studio Reference and his http://www.diyaudio.com/forums/solid-state/162081-dc-servo-mosfet-amplifier.html. you can find almost all of his designs in those three threads.
FH9/FH11 adapted to each other (II); FH9 on FH11 PCB
Until now I provided all files which will be needed for a FH11.
Now you get all information to fit a FH9 on the FH11 PCB.
You only need a adapted schematic and a adapted BoM.
Both you will find attached.
Don't forget to fit the Thiele network or put a jumper!
I will again provide the complete set of files for the FH11 V1.07, this time including the dot marks for the silk screen.....
Until now I provided all files which will be needed for a FH11.
Now you get all information to fit a FH9 on the FH11 PCB.
You only need a adapted schematic and a adapted BoM.
Both you will find attached.
Don't forget to fit the Thiele network or put a jumper!
Attachments
Last edited:
setting bias for Apex ax 14 tef
Hi, mr, i want to try Apex ax14 but i use for load 4 ohm, can i add final transistor again ? And how to set up bias ampere? Thanks for all?
Hi, mr, i want to try Apex ax14 but i use for load 4 ohm, can i add final transistor again ? And how to set up bias ampere? Thanks for all?
The Thiele Network includes both the R+C and the damped L
R24 = 4r7, C24 = 100nF, R27 = 10r and L1 = 2uH
All four components form the Thiele Network.
see my post from January 2007
http://www.diyaudio.com/forums/solid-state/94029-when-output-inductor-needed.html#post1105724
Read the 1975 Dr Cherry article, if you don't believe me.
Last edited:
Many thanks, AndrewT, to clarify the terminology to me! As I could see you deal since many years with this subject. I must confess, I didn't go into details.
Unfortunately I was taken in by information like this:
'There is much confusion about the naming conventions of these output filters. Technically, the RC series network is the Zobel network and LR parallel combination a Thiele network.'
But back to my post:
I made this remark, because without a jumper nothing comes out at the speaker terminal when fitting a FH9. It's not my intention to 'change' in anyway the proved design of FH9 and FH11 (one with and one without the L//R part).
Though my time is limited, I would like to know more about it.
Do you have any link to the Dr Cherry article?
Unfortunately I can get the original Thiele article (I'm not member of the Audio Engineering Society and $33 for information only, is a little too much for me).
But I'm interested in your opinion: Are there good reasons why FH9 and FH11 are different in this output network? And is it true (I read it anywhere) that the values must be adapted to the speaker parameter (probably loudspeaker cable too)?