Symasym - the sequel

Status
Not open for further replies.
First time AMp builder

Hi All
I have decided to tackle a new project and will soon start building a Symasym amp V5.3 using one of AAK's original board design(single pair output). My knowledge of electronics is very basic so I will be asking some pretty basic questions to those with more knowledge than myself(thats alot of people).
I have been hearing alot about transistor matching when reading these forums and would like to know exactly how it is done. I am a visual learner so some diagrams would be great. How much of a difference to an amplifier does transistor matching actually make?
In relation to Symasym which transistors should be matched and should I buy more than the minimum amount needed which is what I bought, to be done properly?

Thanks
Simon
 
Hi Simon
Nothing I can help with unfortunately but this thread hasn't been used for more than a year. Maybe a bit of activity will bring it to someones notice but I think this may not be monitored any more. i.e. no help on this thread. You could PM the principal thread guys and ask, maybe. Check Symasym thread too.
 
Pass website has good info on matching.
hFE function on DMM is almost worthless.
It would do to group a from b from c of the BC series. and subgroup into low and higher.

You need to measure the Ic and Ib at a fixed Vce to determine hFE for that IC and Vce.

You need to compare similar transistors by applying identical Vbe to both while both are at same Tc and both passing very similar Ic This allows you to select pairs of matching Ic at a fixed Vce and fixed Vbe for use in LTP and mirror pairs.
 
Hi Simon,

You need to match Q1&Q2, Q4&Q12, and Q3&Q9. I haven't any problems matching Hfe using my DMM. Andrew's approach is more accurate but since your a beginner it may be to complicated. Buy some extra transistors for matching, normally 2.5 to three times the amount has worked for me. But since the small signal transistors are pretty cheap you may want to buy more just in case you get a mixed batch.

Here's the builders thread that should help with assembly and testing.

http://www.diyaudio.com/forums/solid-state/99739-symasym-5-3-aak-model-builders-thread.html

Regards,

Al
 
Matching Transistor and FETs

If you can build with and embrace SMT, matching the 2N5551, 2N5401 becomes a non-issue. See: PN DMMT5401 and DMMT5551 on Digikey. They are matched pairs, from adjacent die on the wafer. Well matched hfe, Vbe. They are on cut tape, package is SOT-26 (6 lead SOT-23) and the price is OK.

BTW, There are 2 "flavors" of the 2N5551/2N5401. One is 200ma max Ic and the other is 600ma max Ic. I see no functional difference since most circuits are not stressing them that hard to need the 600ma version.

For the 2SK170, getting the SMT LSK389A makes some sense. Common thermal environment and good matching. LTSpice simulates the Symasym (1.3) well with the LSK389A.

My goal, when I build the Symasym 1.3, is to use both SMT and through hole parts when I layout the board. Use SMT for the small signal stuff matched parts in the front end and before the VAS. Use through hole parts for the stuff that gets hot, drivers and outputs. The BD139 will stay as is.

The driving goal of all this matching is to eliminate the evil 'lytic capacitor in series with the feedback resistor to ground and not have a pot to adjust the output offset voltage. I think it can be done with well matched parts.
 
Hi Simon,

You need to match Q1&Q2, Q4&Q12, and Q3&Q9. I haven't any problems matching Hfe using my DMM. Andrew's approach is more accurate but since your a beginner it may be to complicated. Buy some extra transistors for matching, normally 2.5 to three times the amount has worked for me. But since the small signal transistors are pretty cheap you may want to buy more just in case you get a mixed batch.

Here's the builders thread that should help with assembly and testing.

http://www.diyaudio.com/forums/solid-state/99739-symasym-5-3-aak-model-builders-thread.html

Regards,

Al
Thanks for the reply guys. Andrews method sounds a little overboard for a newb like me so I will go with the multimeter method. Also as this board is of AAK's design/build I cannot use the SMT's as suggested but they do seem like an excellent idea as the components are already matched as suggested by Stoc005.
If using hfe method how close do the numbers have to be in regards to the matched components? The following is my results.
Batch of 8 2n5551 hfe 117, 118, 122, 122, 123, 124, 126, 129.
Batch of 6 2n5401 hfe 115, 116, 117, 117, 119, 119
Seems pretty close already? or do I need to buy more?

BTW The AAK's boards are quite valuable as mine has been stuck with the US Postal service for over a week. Just got some movement on the tracking and it's in Philadelphia. I hope it doesn't slow up with Australia Post as well but it wouldn't surprise me (haha, I have great trust in the company I work for).

Thanks
Simon
 
Last edited:
As well as the above post, I have another question as well. I have bought 2 x 25v transformers for 36volt rails and as such is not the same voltage as your AAK BOM which I have worked from. Will I need tochange any component values because I am using 36v rails?
 
it seems from your questions and responses you cannot be bothered to research your choice of project.

Do you really expect us to do your homework?
What do you mean by that? I have already purchased materials for the symasym. I have purchased AAK's board as well. I am posting on this forum as to get some help putting together the amp that is all. Have I insulted you in any way, I do not think so?
 
Hi Simon,

Check your email, I sent you a reply last week that answers your question in post #471. The hfe values you measured are close enough.

Regards,

Al
Thanks for your help Al. I cannot seem to find the email you mention above. I am now schooled up on my Hfe questions though. I did find this quote from the builders thread that you linked.

Hi Sheldon,

For the vbe multiplier you could use MikeB's part selection for R23,R22,R24 and it would work fine with my board, and will allow a wider range of bias control than the values i selected.

The values for the Vbe multiplier that I selected allows for an adjustable range using R22 of between 0mv to about 50mv accross R27 and R28 with +/50V rails. That translates about 0ma to about 62.5ma across each resistor. If you use +/-36V rails you'll need drop R24 to about 240.

If you use MikeB's part values for R23,R22,R24 just make sure that R22 is set to it's lowest bias setting to prevent excessive current on the output devices when you first power up.

__________________
Love when you can, Cry when you have to, be who you must, that's part of the plan.

Al

In the above statement, you said to reduce R24 to about 240ohm if using 36v rails as I will be. Do I need to change your values of R23 (1k) and R22 (100ohm pot) as well? or is it R24 the only resistance change required to suit the 36v supply?

Thanks

Simon
 
all. Have I insulted you in any way,
if your questions/responses were "I do not understand this, please explain" or "I have read this, but when I build it I get the wrong voltages" then we come in to help.

If we hear " I can't be bothered with reading all/any of the previous threads on this project, just tell me what to do"


You expecting us to do your homework is not respecting your fellow Members.

That is an insult.
 
if your questions/responses were "I do not understand this, please explain" or "I have read this, but when I build it I get the wrong voltages" then we come in to help.

If we hear " I can't be bothered with reading all/any of the previous threads on this project, just tell me what to do"


You expecting us to do your homework is not respecting your fellow Members.

That is an insult.

Mate. I am not expecting anyone to do my home work for me I just need some "extra" questions answered as in my post #476 above. I think it shows I have read other posts but need clarification on some points that is all. If someone wishes to answer they can. If they feel "insulted" they do not have to answer. You have already shown your colours already Andrew so I am not really interested in what you have to say anyway. There are many others here who do not show arrogance.

Regards
Simon
 
Hi Simon,

Here's MikeB's SymAsym website.

SymAsym5 - Project

I recommend using the values he selected for R23 (2K), and R24 (500), but for R22 I use a 500 ohm pot instead of a 1K. I've tried this combination from +/-36V to 50+/-V rails and it works great. No other component changes are necessary for +/-36V rails.

Regards,

Al
 
Status
Not open for further replies.