The (semi) Big FET shoot-out

[crbowman]

Just tried out a few different FETs in my 70's style amp. My impressions are simply opinions and YMMV. Gain trim pot was left at it's mid-way point for a more unbiased observation.

2N5457 (Fairchild): Works perfectly fine, but a little low in gain comparatively. Also a little thin and sort of sterile, but was slightly less noisy than the others listed (which were all relatively the same as far as noise level). Personally, this was my least favorite. Mouser part # 512-2N5457

2N5486 (Central Semi): Compared to the 5457, fatter, warmer, more sustain (more gain), and seemed more 'musical'. Notes bloomed into sustained feedback much more easily at the same setting of the trim pot, and cleans were big open and complex. This one was my favorite. Mouser part # 610-2N5486

2N4416A (Central Semi): This one is a 4 lead FET and I only have a 3 pin socket installed. Left the case lead unconnected so take this for what it's worth. Sounded similar to the above 2N5486 Big open cleans, very fat and warm. Something a little weird occasionally in the upper mids in overdirve. Hard to describe, kind of a slight fizz. Only seemed to happen on certain notes, and then not all the time. All in all, still sounded better to me than the 5457. Mouser part # 610-2N4416-A

2N5486 (Fairchild): Almost identical to the Central Semi version. Could be my imagination but the Central Semi version just seemed slightly more 'musical' somehow. Hard to describe, but there. Mouser part # 512-2N5486

By all means this is all IMHO, and in my amp. YMMV.
Hope this helps anyone out there trying to decide on an FET.
Peace,
CB

[selloutrr]

Thank you!

[talbany]

crbowman

Looking for a FET for my newly built 40 .. Gracias!!


Tony

[David Root]

Very helpful, thank you! Do you have an NTE452 you could compare?

[crbowman]

Very helpful, thank you! Do you have an NTE452 you could compare?

Sure don't, sorry David. Wish I did, but I was surprised at what they're asking for them now. This is more like the 'Bargain Basement" shootout.

[crbowman]

crbowman

Looking for a FET for my newly built 40 .. Gracias!!


Tony

De nada!
Glad to help Tony.
I'm interested in your impressions of this amp when you're done.
I'm really liking mine.

[markr14850]

Did you adjust the bias for each FET?

Also, did you measure Idss or Vp for any of them? These can vary wildly, even in the same batch. A lower Vp will generally give more gain - but will clip sooner. Were you running single coils or humbuckers?

--Mark

[JD0x0]

Did you adjust the bias for each FET?

Also, did you measure Idss or Vp for any of them? These can vary wildly, even in the same batch. A lower Vp will generally give more gain - but will clip sooner. Were you running single coils or humbuckers?

--Mark

He mentioned in his post that he kept it about halfway and did not adjust the bias to provide the most controlled results.

[crbowman]

Did you adjust the bias for each FET?

Also, did you measure Idss or Vp for any of them? These can vary wildly, even in the same batch. A lower Vp will generally give more gain - but will clip sooner. Were you running single coils or humbuckers?

--Mark

I have a 10K trim pot in place of the 8.8K resistor. Kept it near the top of the rotation. Did no real measurements though. I was using my '69 Gretsch Tennessean with the stock Hi-Lo Trons which are pretty low output single coils. Not what most people use here, I know, but it's my baby. It's also the main reason I am so interested in getting the FET input sounding it's best.

[markr14850]

He mentioned in his post that he kept it about halfway and did not adjust the bias to provide the most controlled results.

The bias is set by the source resistor (the FET equivalent of a cathode resistor), not the output level trimmer.

I think it's important to adjust the bias when trying different FETs (even with the same part number). Think about it this way - you wouldn't swap 6V6, 6L6, EL34, and KT88 tubes into an amp for comparison without *at least* adjusting the bias.

Also, leaving the output trimmer at half isn't a fair comparison, as different FETs will have different gain - and more gain is quite often interpreted as "better". It would be a more accurate comparison to use a signal generator and a scope and set the trimmer so as to always give the same output level.

Just my opinion,
--Mark

[markr14850]

I have a 10K trim pot in place of the 8.8K resistor. Kept it near the top of the rotation. Did no real measurements though. I was using my '69 Gretsch Tennessean with the stock Hi-Lo Trons which are pretty low output single coils. Not what most people use here, I know, but it's my baby. It's also the main reason I am so interested in getting the FET input sounding it's best.

I hear you there. I send my strat into the FET input and my HB into normal. I set the FET output level trimmer to make them have roughly the same level.

I don't usually send HB's into my FET. Depending on the Vp of the FET, HB's may be too hot, and can clip hard. I don't think that this will happen with single coils for any of the usual FET recommendations.

For the bias, a reasonable way is to adjust your source resistor (trimmer) so that the voltage at the FET drain is half of the supply voltage. That probably works out to about 10V. My guess is that this will be closer to 1kohm than to 8.8k.

BTW, here's my tweaked version of the FET circuit: http://ampgarage.com/forum/viewtopic.php?t=9561

Enjoy,
--Mark

[crbowman]

Very cool!
Nice way of compensating for SC or HBs. I'll have to check out your FET circut mods when I get around to the tweaking stage. I actually did adjust the bias trimmer by ear when I installed the first FET, but I must admit I didn't really mess with it much beyond that point.
Thanks Mark