# Millett "Starving Student" hybrid amp

Tags:

1. Quote:

 Originally Posted by the_equalizer /img/forum/go_quote.gif Another thing I don't yet understand: does the source-gate voltage have to be around 5 volts ? Can it be higher? How much higher? Could the 12AU7 version work with the same voltage divider as the 19J6 version (e.g. with the gate sitting at 24 volts with respect to ground) ? I can see that this voltage (along with the source-drain voltage) determines how much voltage the MOSFET can swing, but I haven't yet sat down to carefully read the datasheet and run some experiments with the amp

Vgs is a property of the Mosfet and changes depending upon how much current is going through it. That is, there is a correlational relationship between Ids and Vgs. This can work in two ways -- first, as here, as the current increases, Vgs also increases, and as the current decreases, Vgs decreases. Vgs is pretty much always close to 4.5V for the IRF510 and IRF610, and just moves up a small amount with different currents -- too small an amount for concern on this amp.

However, you can use it the other way, too. That is, by carefully controlling Vgs, you can control the amount of current that goes through the mosfet thus making it a constant current source. So, for instance, you can use a resistor divider from + to bias the gate. Then, by varying a resistor between the source and ground, a resistor that drops voltage across it, you can control the voltage between the gate and the source, which controls the amount of current that can flow.

So, in the Starving Student, why is it not acting like a CCS? The reason is because the mosfet is biased into saturation. The amount of voltage we have between the gate and the source is too high to be able to control the current, so the mosfet will let basically any amount of current flow. Instead, the current is limited by the heater which, at the bias voltage of the mosfet minus Vgs (~19V), can only draw 150mA thus limiting the current across the mosfet. You can think of the heater as the resistor used with an LED -- the LED will pass full current and burn itself up without a current limiting resistor inline with it. Here, the heater acts as as the limit. The mosfet differs from the LED, though, in that it has an extra leg whose voltage with regard to the cathode is set by the current between the anode and the cathode.

2. Quote:

 Originally Posted by the_equalizer /img/forum/go_quote.gif I'm not sure I quite follow you there with the "double the mosfets" idea.

unless I'm wrong, you should be able to run the mosfets in parallel just to dissipate the heat between two mosfets instead of one.

3. Quote:

 Originally Posted by dsavitsk /img/forum/go_quote.gif Vgs is a property of the Mosfet and changes depending upon how much current is going through it. That is, there is a correlational relationship between Ids and Vgs. This can work in two ways -- first, as here, as the current increases, Vgs also increases, and as the current decreases, Vgs decreases. Vgs is pretty much always close to 4.5V for the IRF510 and IRF610, and just moves up a small amount with different currents -- too small an amount for concern on this amp. However, you can use it the other way, too. That is, by carefully controlling Vgs, you can control the amount of current that goes through the mosfet thus making it a constant current source. So, for instance, you can use a resistor divider from + to bias the gate. Then, by varying a resistor between the source and ground, a resistor that drops voltage across it, you can control the voltage between the gate and the source, which controls the amount of current that can flow. So, in the Starving Student, why is it not acting like a CCS? The reason is because the mosfet is biased into saturation. The amount of voltage we have between the gate and the source is too high to be able to control the current, so the mosfet will let basically any amount of current flow. Instead, the current is limited by the heater which, at the bias voltage of the mosfet minus Vgs (~19V), can only draw 150mA thus limiting the current across the mosfet. You can think of the heater as the resistor used with an LED -- the LED will pass full current and burn itself up without a current limiting resistor inline with it. Here, the heater acts as as the limit. The mosfet differs from the LED, though, in that it has an extra leg whose voltage with regard to the cathode is set by the current between the anode and the cathode.

Thanks for that! I'm already reading up on biasing MOSFETs into their saturation region.

I'd have just one more question, with Vgs sitting at 4.5 volts, does that mean that the maximum input signal amplitude the MOSFET can take without being driven out of saturation (clipping) would be around 9 V p-p ?

Thanks again!

Quote:

 Originally Posted by nullstring /img/forum/go_quote.gif unless I'm wrong, you should be able to run the mosfets in parallel just to dissipate the heat between two mosfets instead of one.

Ok, I see what you mean. Well, it looks as if it could work, yes, some kind of parallel source follower. You'd also have to include a gate resistor (R3a, let's call it) for the new MOSFET: one end connected directly to the MOSFET's pin 1 and the other end connected to the junction of R2, R4 and R3; this would bias the MOSFET to conduct and input the signal into it's gate.

Still, it seems to me that the space occupied by the new MOSFET and it's heatsink could as easily be used to harbour a BIG heatsink for a lone MOSFET, plus you wouldn't need another heatsink mounting kit.

On the other hand the four heatsinked MOSFETs would look cool . Time to build a prototype maybe ?

cheers!

4. double-post

5. Yep, 48V and anything over .350A is good. If you're doing one of the mod versions you might want more current though, I seem to remember that's called for but I can't find numbers.

6. I'm doing the 12AU7 Version. According to this schematic, the same source can be used. Is this just an oversight on the_equalizer's part?

Also, am I right in thinking that just one of these can be used to cool both mosfets? I'm looking for any means of cutting down the price on this bad boy.

While I'm asking questions, how do you go about mounting this to a chassis? Seems like it'd be pretty straight-forward but the picture is not all that great.

7. Quote:

 Originally Posted by ezzieyguywuf /img/forum/go_quote.gif Cisco Power Supply PN 34-1977-03 model PSA18U-480C - eBay (item 320483355045 end time Apr-03-10 16:02:27 PDT) theres an example. Says 0.38 amps. I can't get that one cuz it doesn't have the power cable :-/ otherwise it's an awesome deal at \$5.

You know that the cable from main to the power supply is just standard computer cable, right?

I don't know the proper name for them, but it's just the cable that tends to go from 120VAC to a desktop computer PSU.
The type that tends to DIY audio, tends to build their own computers, so it wouldn't be uncommon if you already had a few extras.

Although, now that I look at the link, the shipping is \$16.. which kills the deal anyway, making this post useless.. oh well, I already typed it.

8. I'm doing the 12AU7 Version. According to this schematic, the same source can be used. Is this just an oversight on the_equalizer's part?

Also, am I right in thinking that just one of these can be used to cool both mosfets? I'm looking for any means of cutting down the price on this bad boy.

While I'm asking questions, how do you go about mounting this to a chassis? Seems like it'd be pretty straight-forward but the picture is not all that great.

<edit>
Quote:

 You know that the cable from main to the power supply is just standard computer cable, right? I don't know the proper name for them, but it's just the cable that tends to go from 120VAC to a desktop computer PSU. Although, now that I look at the link, the shipping is \$16.. which kills the deal anyway, making this post useless.. oh well, I already typed it.

I found this deal, which is \$10 shipped, so I'll just get that one. Thanks though.

9. Quote:

 Originally Posted by ezzieyguywuf /img/forum/go_quote.gif I'm doing the 12AU7 Version. According to this schematic, the same source can be used. Is this just an oversight on the_equalizer's part?

No, it looks ok, I'm just getting things messed up in my head, that's all. Nice to know the power adapters are still readily available.

10. Ah I still have yet another question actually: when wiring up the mosfets, should I use shrink wrap or something to keep the feet from touching each other? Is there a standard way of doing this? Also, I'm guessing my enclosure will have to have a hold for the mosfet wires to go through?

11. Quote:

 Originally Posted by ezzieyguywuf /img/forum/go_quote.gif Ah I still have yet another question actually: when wiring up the mosfets, should I use shrink wrap or something to keep the feet from touching each other? Is there a standard way of doing this?

It depends ... remember that the MOSFETs are going to be bolted down, so there's not much chance of them moving. I'd still do it, but then I'm the cautious type.
Quote:

 Also, I'm guessing my enclosure will have to have a hold for the mosfet wires to go through?

Most people used rubber grommets to insulate the edges of a hole from contacting the MOSFET leads. Look back at Pete's original, Nate's, and some of the early P2P builds.

12. thanks tomb, I'll look at those earlier builds.

I'm beginning to think more and more that enclosing this thing is going to be infinitely worse than getting the circuit together. My plan is actually to build the circuit next week, and then maybe wait till the summer to design and build a case for it darn grad school is sooo time consuming. Only thing is, I have access to a machine shop now (through one of my classes), and I won't later in the summer. *sigh*.

13. Koss PortaPro + SSMH = match made in heaven

that is all

14. Contributor
Quote:

 Originally Posted by ezzieyguywuf /img/forum/go_quote.gif I'm beginning to think more and more that enclosing this thing is going to be infinitely worse than getting the circuit together.

You catch on quick.

Casework can very easily make up most of the time and money spent on a DIY project.