I know that this is a newbie question, but where are the test points for the offset? Is it the positive and negative speaker outputs and the adjustment via R105?
I think I'll go for the lower voltage version since my speakers don't need much voltage but rather some current.
Some ideas:
to get asymmetric rails I could use a regular transformer with symmetric outputs, but only add on the positive rail a Juma capacitance multiplier.
That will drop the positive rail with 2v.
For mosfets something from Ixys like the 40p50p 40n30p in TO264 package.
Those are also used in Pass XA25 but in SOT227.
Other suggestions ?
Are there any reliable spice models for those new mosfets ?
So far I've used the cordell spice lib in ltspice.
Regards,
Danny
You could use 4x fatty heatsinked TO220 diodes in series to drop 2.4v.
Hugh uses triple IRFP240/9240 to simulate a big IXYS. Basically have same three times the input capacitance.
Hugh uses triple IRFP240/9240 to simulate a big IXYS. Basically have same three times the input capacitance.
Danny,
That would work. Decrease the rail voltage to around 18V and you can increase the quiescent to 2.47A with 0.27R sense resistor.
Then it will go easily down to a 2.5R load. The large Fairchild or IXYS 40N25 rated to 280W would be just fine.
Cheers,
Hugh
That would work. Decrease the rail voltage to around 18V and you can increase the quiescent to 2.47A with 0.27R sense resistor.
Then it will go easily down to a 2.5R load. The large Fairchild or IXYS 40N25 rated to 280W would be just fine.
Cheers,
Hugh
Does that mean for R105 you don’t need the variable resistor, but can use a fixed resistor? If so what value or what should I set the variable to?
Yes, you do not use the variable resistor, it's not a particularly good idea. Specify the current you want and then set the resistor on the basis of 0.62/Is. I suggested 0.27R for 2.3A, plus or minus a few percent according to the sense transistor's Vbe which I'm assuming around 0.62V.
Danny, I have attached my latest of the low voltage version set up for high current/low impedance drive. (Ignore some of the text I have put in for the 8R instance). The negative and positive clip is more six volts off rail, around -11.2V and +11.64, but this would improve considerably with high power high transconductance mosfets, which do not require so much Vgs to push through high current. I would not suggest using asymmetric rails; the power output into 3R at -11.1V and +11.1V, just below clip, is 20.5W, into 4R it's -11.4V and +15.0V, just below clip and a lower 15.4W. In either case the distortion figures are very good, typically 0.05% with H2 -69dB and H3 -72dB. Note that in these cases the third harmonic comes up and even exceeds the second at full output; this is a function of low transconductance devices in the output stage and I'd strongly suggest using the Fairchild or IXYS devices rated at 480W here since the increase of Vgs and high current is minimised compared to the IRFs products.
Hugh
Danny, I have attached my latest of the low voltage version set up for high current/low impedance drive. (Ignore some of the text I have put in for the 8R instance). The negative and positive clip is more six volts off rail, around -11.2V and +11.64, but this would improve considerably with high power high transconductance mosfets, which do not require so much Vgs to push through high current. I would not suggest using asymmetric rails; the power output into 3R at -11.1V and +11.1V, just below clip, is 20.5W, into 4R it's -11.4V and +15.0V, just below clip and a lower 15.4W. In either case the distortion figures are very good, typically 0.05% with H2 -69dB and H3 -72dB. Note that in these cases the third harmonic comes up and even exceeds the second at full output; this is a function of low transconductance devices in the output stage and I'd strongly suggest using the Fairchild or IXYS devices rated at 480W here since the increase of Vgs and high current is minimised compared to the IRFs products.
Hugh
xrk971,
I will do that.
Hugh,
I do not have your level of knowledge, so I’m not tracking. Are you saying I should make another change? Where would that be?
Brad
Hi BRN,
Hugh was actually talking about Danny_66’s low impedance amp request with 2.3amp bias that is stable into 2.7ohm load.
For Alpha 20, I am confident now it needs no pots at all. Just make R115 a 1050 ohm resistor. Probably 1000+47R is fine.
Hugh was actually talking about Danny_66’s low impedance amp request with 2.3amp bias that is stable into 2.7ohm load.
For Alpha 20, I am confident now it needs no pots at all. Just make R115 a 1050 ohm resistor. Probably 1000+47R is fine.
Thanks X,
You are 100% correct here. I think on reflection that the stage current should be set at with 3k3, then 6k8, then 1k as you suggest. If you are shy about 137R, the correct figure, the change in the offset is only about 3mV down, which is trivial.
Put another way, with a setting of 3k3/6.8k/1k, we have an offset of -3.35mV.
Set at 3k3/6.8k/1k2, the offset is +1.52mV, 200R leads to about 4.87mV, around 100R gives 2.4mV.
In truth, even for a Lowther an offset of up to 25mV is not important. It won't damage the voice coil, it won't change the sound, but it should be thermally stable, and this is where the LTP and Lender drive with diode pays dividends.
Cheers,
Hugh
You are 100% correct here. I think on reflection that the stage current should be set at with 3k3, then 6k8, then 1k as you suggest. If you are shy about 137R, the correct figure, the change in the offset is only about 3mV down, which is trivial.
Put another way, with a setting of 3k3/6.8k/1k, we have an offset of -3.35mV.
Set at 3k3/6.8k/1k2, the offset is +1.52mV, 200R leads to about 4.87mV, around 100R gives 2.4mV.
In truth, even for a Lowther an offset of up to 25mV is not important. It won't damage the voice coil, it won't change the sound, but it should be thermally stable, and this is where the LTP and Lender drive with diode pays dividends.
Cheers,
Hugh
I have capMx but there is a small thumb, actually it's not thumb but same kinda cone movement that Vunce has also. Don't want to risk breaking 2" driver though, since I drive it through DSP without passive filtering.
When using a DSP, it is always good practice to add a capacitor in series with your tweeter, even if your amp has no turn on and turn off transients.
Hi Hugh,
Thanks for the high current/low impedance simulation !
Very nice that you also got low distortion for 3r and 4r,
this will be lower with the typical output of -2v+2v that I only need to play already loud (94dbc)
I'll definitely use Fairchild or IXYS devices rated at 480W.
In my current F5 I also use Fairchilds.
Regards,
Danny