I wonder if Hugh can chime in and comment on the ‘slam and impact’ part with regards to the Darlington in this circuit? I am wondering what technically is happening and if it has to do with changes in transconductance or if it is because it’s a BJT characteristic?
https://www.digikey.com/en/products/detail/sanken-electric-usa-inc/2SD2560/3661819
https://www.semicon.sanken-ele.co.jp/sk_content/2sd2560_ds_en.pdf
I’ll be playing with the power JFET discussed earlier in this thread and can add this BJT Darlington into the comparison mix
Best,
Anand.
https://www.digikey.com/en/products/detail/sanken-electric-usa-inc/2SD2560/3661819
https://www.semicon.sanken-ele.co.jp/sk_content/2sd2560_ds_en.pdf
I’ll be playing with the power JFET discussed earlier in this thread and can add this BJT Darlington into the comparison mix
Best,
Anand.
Last edited:
Have just done a 'clipping' measurement with a 4 ohm load.
p-t-p voltage at just before clipping is 38v (with +/-21v rails) so - if my calculations are correct - this gives 45w into 4 ohms:
* 38v p-t-p ==> 19v peak
* ==> 13.4v rms
* power = Vrms^2 / load
* which is 180 / 4
* ==> 45w.
This is the same for my other AN4R - which hasn't yet had its Source resistors changed from 0.12R to 0.15R.
Which shows that the bias current doesn't affect the power output - so I guess it only affects:
* how well the amp drives low loads
* and of course ... the heat output into the room!
p-t-p voltage at just before clipping is 38v (with +/-21v rails) so - if my calculations are correct - this gives 45w into 4 ohms:
* 38v p-t-p ==> 19v peak
* ==> 13.4v rms
* power = Vrms^2 / load
* which is 180 / 4
* ==> 45w.
This is the same for my other AN4R - which hasn't yet had its Source resistors changed from 0.12R to 0.15R.
Which shows that the bias current doesn't affect the power output - so I guess it only affects:
* how well the amp drives low loads
* and of course ... the heat output into the room!
Τhanks Andy for your measurements.
The difference of bias has impact to quality of sound? Do you heard any changes?
Haven't listened to them yet, Nikos. (I need to wait until I have changed over R141/142 in my 2nd AN4R, to be able to do a proper listen.)
But I suspect I won't be able to make a valid comparison ... as it's so long since I listened to the amps in their initial state!
But we do know that higher bias current enables the AN circuit to drive lower spkr loads, well. Somewhere in these 154 pages of posts is a report from someone who tried to drive 4 ohm spkrs with his AN8R (1.7a of bias current) ... he said it didn't sound good.
Whereas my AN4R circuit - with 3.2a bias - drives 4 ohm spkrs very well.I have to wait and see whether reducing the bias current to 2.5a (by increasing R141/142 from 0.12ohm to 0.15ohm) has degraded the sound with my 4ohm spkrs.
It has certainly reduced the heat output into the room - down from 500w to 420w, for 4 channels.So I'm hoping I get another couple of months in the year when I am able to run them.
Probably not me:
https://www.diyaudio.com/community/...8ohm-class-a-amp.344540/page-139#post-7496289
https://www.diyaudio.com/community/...8ohm-class-a-amp.344540/page-122#post-7320024
Thanks for your recent measurements Andy!
I also found a post by Hugh which is relevant with regards to 4 ohm loads:
Best,
Anand.
Last edited:
Anand, Andy, Nikos and Vunce:
What to expect with a darlington bipolar in place of the nmos on the AN39?
Well, as Andy has found, 'slam and impact' probably explains the effect reasonably correctly. But why?
Mosfets gates are not galvanically attached to their source but bases on bipolars are directly connected to the emitters.
Any capacitive, gate connection is an indirect connection, so you cannot expect instantaneous response to inputs, unlike bipolar transistors.
Transistors do have high transconductance, around 25S, but large mosfets are high too. But source follower devices drive speakers through source resistors, and same for transistors. The gm of a source/emitter follower is dominated by the value of the source/emitter resistors so I don't believe gm of the individual device is the issue between the sonic differences.
I'm at a loss to explain in more detail, but I can give a useful analogy. Mosfets operate like gasoline engines, bipolars like diesels. Anyone who has owned both cars would characterise a diesel engine as offering 'slam and impact', whilst a gasoline motor is better described as 'refined and high speed response'. This explains the use of diesels for heavy traction purposes, where slam and impact is needed, whereas gasoline engines are ideal for acceleration and high speed.
My apologies, slow response, I have been distracted by many other issues, including a DHT preamp I'm designing.
Cheers,
Hugh
What to expect with a darlington bipolar in place of the nmos on the AN39?
Well, as Andy has found, 'slam and impact' probably explains the effect reasonably correctly. But why?
Mosfets gates are not galvanically attached to their source but bases on bipolars are directly connected to the emitters.
Any capacitive, gate connection is an indirect connection, so you cannot expect instantaneous response to inputs, unlike bipolar transistors.
Transistors do have high transconductance, around 25S, but large mosfets are high too. But source follower devices drive speakers through source resistors, and same for transistors. The gm of a source/emitter follower is dominated by the value of the source/emitter resistors so I don't believe gm of the individual device is the issue between the sonic differences.
I'm at a loss to explain in more detail, but I can give a useful analogy. Mosfets operate like gasoline engines, bipolars like diesels. Anyone who has owned both cars would characterise a diesel engine as offering 'slam and impact', whilst a gasoline motor is better described as 'refined and high speed response'. This explains the use of diesels for heavy traction purposes, where slam and impact is needed, whereas gasoline engines are ideal for acceleration and high speed.
My apologies, slow response, I have been distracted by many other issues, including a DHT preamp I'm designing.
Cheers,
Hugh
Last edited:
Thanks for the explanation, Hugh.
DHT preamp, Hugh ... sacrelige!
Sand rules!Never having owned a diesel car (IMO ... a diesel Golf is an abomination!), I would've thought the two engines are better compared as:
* a diesel engine has grunt & torque ... to pull heavy loads
* whereas a petrol engine has a more instantaneous impact with a light vehicle ... when it can respond better to the accelerator.
So Mosfets probably cope with low spkrs loads better than a BJT ... but BJTs deliver much better transients?
I would have thought bipolars outputs give you 'slam and impact', and arguably better transients?
I must have expressed this badly......... but the fact is bipolars do have SOAR limitations so one must be careful with low impedance loads.
They do it very well with suitable design; for 200W that needs at least three pairs of BJTs, whereas a single, large pair of mosfets can do the same power in a low impedance load, like the Maya.
Hugh
I must have expressed this badly......... but the fact is bipolars do have SOAR limitations so one must be careful with low impedance loads.
They do it very well with suitable design; for 200W that needs at least three pairs of BJTs, whereas a single, large pair of mosfets can do the same power in a low impedance load, like the Maya.
Hugh
Hey all,
I’m getting ready to pull the trigger on the parts for the an39 8ohm. Have a few questions:
1) I’m going to use the mouser cart BOM. Is there anything I need to know before pulling the trigger on it? I did see a few parts that were not in stock and end of life. Are these transistors good to use FQA40N25?
2) Also, I’m using the SLB board. Does anyone know if there is a mouser cart file for this? is there anything I should know about buying those parts?
3) Lastly, I’m also using the soft start board. Same questions as above?
Thnx,
🙂
I’m getting ready to pull the trigger on the parts for the an39 8ohm. Have a few questions:
1) I’m going to use the mouser cart BOM. Is there anything I need to know before pulling the trigger on it? I did see a few parts that were not in stock and end of life. Are these transistors good to use FQA40N25?
2) Also, I’m using the SLB board. Does anyone know if there is a mouser cart file for this? is there anything I should know about buying those parts?
3) Lastly, I’m also using the soft start board. Same questions as above?
Thnx,
🙂
BMM,
1) Please tell us which transistors are ‘not in stock.’ There are likely substitutes.
2) SLB Mouser BOM link is here. Note you can use a different LED for the LTL4321N or substitute or not populate the LED since it doesn’t serve any function other than an “indicator.” Questions for the SLB should directed to that thread just to keep things organized. Try these substitutes:
https://www.mouser.com/ProductDetail/Lite-On/LTL-4231N-1?qs=A6JmxJ40G%2BbitDfMpylDkg==
https://www.mouser.com/ProductDetail/Lite-On/LTL-4231?qs=ijkchSGtXvugYEuZKOTKWw==
3) No idea about the softstart board. Try here. Also be VERY specific about which softstart board/version you are looking for. Things have changed on the softstart board since its first inception.
Best,
Anand.
1) Please tell us which transistors are ‘not in stock.’ There are likely substitutes.
2) SLB Mouser BOM link is here. Note you can use a different LED for the LTL4321N or substitute or not populate the LED since it doesn’t serve any function other than an “indicator.” Questions for the SLB should directed to that thread just to keep things organized. Try these substitutes:
https://www.mouser.com/ProductDetail/Lite-On/LTL-4231N-1?qs=A6JmxJ40G%2BbitDfMpylDkg==
https://www.mouser.com/ProductDetail/Lite-On/LTL-4231?qs=ijkchSGtXvugYEuZKOTKWw==
3) No idea about the softstart board. Try here. Also be VERY specific about which softstart board/version you are looking for. Things have changed on the softstart board since its first inception.
Best,
Anand.
Last edited: