Ive just made a headphone amp version for my Sennheiser HD 800's. 15V at 50ma bias. I simplified the the upper CFP to use a BC337 and an MJE15031 while the lower current source uses
another BC 337 and a MJE15030. Over all gain reduced to 4.5. Its powered with 4 x 18650 Lithium Ion batteries for now. First impressions are that it sounds very very good. I think it could be even better than my DHT valve headphone amp which was a big surprise. Now to do some comparable measurements.
another BC 337 and a MJE15030. Over all gain reduced to 4.5. Its powered with 4 x 18650 Lithium Ion batteries for now. First impressions are that it sounds very very good. I think it could be even better than my DHT valve headphone amp which was a big surprise. Now to do some comparable measurements.
Would love to see them!Most people say something like used DC-power *2 for Class-A.
So the original circuit has 20V*1A*2channel=40W and thus account for at least 2*40=80VA.
You can try the circuit with anything from 16 to 24V if you have a laptop power supply around or similar.
(same here though: Supply should have at least 1.5-2x the current rating you actually need)
Great, thanks! I have plenty of 19v high power laptop power supplies and will keep them in mind. Ideally I will be able to squeeze a transformer and small CRC PSU in the case, the fall back plan if space doesn't allow is the laptop supplies.
dliscomb
I did not fully explain the MoFo situation. Yes, using an inductor would give you a MoFo SE amplifier almost doubling efficiency over using a simple CCS.A single ended amp processes the entire waveform, positive and negative, and the CCS used on this Class A 4W amp sinks current for the negative halfcycle, with output device sourcing the current for the positive halfcycle.
You can replace the CCS with a passive device, an inductor. It's big, expensive and heavy but it has some advantages.
An inductor, which stores energy in a magnetic field, can sinks current for the negative halfcycle like the CCS by Lenz' Law, whereby the output of the inductor reacts to any break or change in current. So, if you run 1A through it, the magnetic field will react by changing voltage to attempt to maintain that current should that current be altered. You set the current through the inductor statically by using its DCR to fix a current at its input.
For a DCR 0.5R inductor, then 500mV setting at the output of the inductor will ensure 1A passes through it to ground. Now, if the audio signal goes negative at the gate, the inductor will maintain its current by creating a negative voltage at its output. This voltage is actually set by the mosfet; about 4.5V lower than the gate voltage. Since the inductor has stored energy, it will go negative until the stored energy is spent. This will supply current and voltage set by the load and the source voltage. The inductor in this way obliges with a negative voltage at the source to follow the gate, and you can create almost as much negative voltage the mosfet source/top of the inductor as the positive voltage the mosfet passes positive halfcycle voltage onto to the load from the Vcc rail. There are losses of course but with a good inductor you should reach much of the Vcc. If you are using 18V, remove the Vgs your positive halfcycle could reach about 13V less the inductor static drop, of say 1V. So that would be 12V positive halfcycle, and maybe as much as 10V negative halfcycle, so you should be able to generate almost 22Vpp into a load.
If you have an inductor at DCR 0.5, the Vcc is 18V and you are setting it up for an 8R load with a quiescent of 1.5A, you could get about 7.5W audio with a total input of 18 x 1.5 = 27W. This is 28%, not too shabby for Class A. Most of the dissipation would be taken by the single mosfet.
HD
Last edited:
Hi
anybody know about the 5200 version of UTC?
because the toshiba are often not available
UTC SC5200
anybody know about the 5200 version of UTC?
because the toshiba are often not available
UTC SC5200
I powered up one channel today using a 24vdc 5a laptop adapter. Used a 100R/2W resistor between the output and ground as Juma suggested. Because I had 1R/3W resistors so used a pair and then the rest of the parts are exactly per the schematic. 2SC5200 pairs as power mosfets. I see the bias across one of the 1R resistor being 615mV with the offset very close to 0.5mV and stable. The trimmer does not change the bias or reduce not sure if the trimmer is the issue or I used a slightly lower value 1R power resistors parallel.
Thanks
Thanks
@AKSA
Hugh,
Many thanks for that excellent follow-up; as @Vunce said, there's much for me to take in.
ls there a point in time when the steady flow of current through the inductor causes core grain alignment enough that it would need to be replaced? I'm thinking there is a steady DC current from the bias on the upper transistor, no?
Or is the AC nature of the signal enough to keep it nonpolar? Assuming we don't leave the amp running without signal for a long time.
I understand electrolytic capacitor degredation (enough), but I hadn't thought about inductor "degredation" until now.
Kind regards,
Drew
Hugh,
Many thanks for that excellent follow-up; as @Vunce said, there's much for me to take in.
ls there a point in time when the steady flow of current through the inductor causes core grain alignment enough that it would need to be replaced? I'm thinking there is a steady DC current from the bias on the upper transistor, no?
Or is the AC nature of the signal enough to keep it nonpolar? Assuming we don't leave the amp running without signal for a long time.
I understand electrolytic capacitor degredation (enough), but I hadn't thought about inductor "degredation" until now.
Kind regards,
Drew
Its playing some beautiful music in a single channelOk I read the post #4 again and as @lineup suggested I removed the 100R/2w output resistor and was able to adjust the DC output voltage to V/2 + 0.8 so my supply is 24/2 + 0.8 to set it to 12.8vdc using the trimmer
Thanks
Plays as good as ACA if I can say very good mids and vocals the forte of class A. Thanks @lineup for a sweet sounding amplifier and can be built with super cheap parts without any exotic JFETs or mosfets.
Last edited:
The power inductors made for this purpose don’t “wear out”. They may become partially magnetized but that doesn’t affect their ability to act like an inductor and provide the reactance to nearly double the output swing. Look at the Mofo thread or LuFo threads for examples. Microwave oven transformers are great for this. North American ones (120 vac) are circa 0.5ohms and 70mH and $30 on eBay. EU and rest of world where 220vac is norm don’t work as well as they are 1ohm+ DCR. Too high.
For LuFo amp, 28v single rail supply (4v more than here) allows amp to make 39w into 8ohms through the miracle of the inductive load instead of a CCS.
Could you share a link or part number to the sockets you used for R6?
I see a lot of options out there and I just want to make sure I get the right thing. Thanks!
I'm thinking of giving this a go as my first build so excuse any newbie questions. My speakers are very efficient so this is ideal - I'm listening in the milliwatts region
Are the 2sc5200 output transistors equivalent with MJL4281?
Would a soft-start module be essential given the sensitivity of my speakers and if so what's a good option to look at?
Thanks
Are the 2sc5200 output transistors equivalent with MJL4281?
Would a soft-start module be essential given the sensitivity of my speakers and if so what's a good option to look at?
Thanks