diyAB Amp The "Honey Badger" build thread

[danielwritesbac]

I am terribly confused by Antek - AN-4235 with the test conditions stating secondaries were parallel during the test. Normal hookup would be with the secondaries effectively series, but the problem is that the datasheet doesn't give any specs for using the transformer normally. This is a maximized core, minimized copper build (a giant core barely covered in thin gauge copper), so I assume that secondaries series normal use would perform a bit differently than their secondaries parallel test.
Question:
If I would like the equivalent of an old fashioned 5.7A 70VCT (35+35VAC Center Tap), to power each monobloc, then would I need Antek - AN-6238 transformers?

[6L6]

There is no doubt that the Antek spec sheets cause some confusion. Oh well. At least they have some measured data, many manufacturers or distributors don't even have that.

Something to keep in mind - in the description they state that " In most of the cases, this transformer can be output 20% more power from its rating at 60Hz power source without any problem."

If that isn't enough to settle your worries, then just buy the next bigger one.

[Tony]

Originally Posted by danielwritesbac

I am terribly confused by Antek - AN-4235 with the test conditions stating secondaries were parallel during the test. Normal hookup would be with the secondaries effectively series, but the problem is that the datasheet doesn't give any specs for using the transformer normally. This is a maximized core, minimized copper build (a giant core barely covered in thin gauge copper), so I assume that secondaries series normal use would perform a bit differently than their secondaries parallel test.
Question:
If I would like the equivalent of an old fashioned 5.7A 70VCT (35+35VAC Center Tap), to power each monobloc, then would I need Antek - AN-6238 transformers?

you mean short circuit tests? the object of the test is to determine copper losses and so the testing calls for all secondaries of the "traffo under test" to be effectively shorted together, rated primary current is then allowed to flow using a variac to set a voltage that allows this....this procedure is common to all transformers undergoing such test...

add the copper losses(short circuit) and the core losses(no-load test) and you can get the efficiency of the traffo....to the average user, this is of no consequence....

a 600VA traffo for the honey badger is more than enough imho....

[6L6]

Originally Posted by danielwritesbac

Question:
If I would like the equivalent of an old fashioned 5.7A 70VCT (35+35VAC Center Tap), to power each monobloc, then would I need Antek - AN-6238 transformers?

Ok, you lost me.

5.7A at 70vct is 400VA An Antek AN-4235 would be exactly what you are asking for.

38v+38v (or 76vct) at 600VA is 7.9A
(And even better, the Antek 6238 says it can flow 2x 38v windings at a total of 9.4A -- that's more than 700VA!)

[Tony]

a power traffo rated for 7.9 amps does not stop delivering currents at 8amps....

Va ratings are all about tolerable heating that can be allowed in a traffo....

[danielwritesbac]

Originally Posted by 6L6

5.7A at 70vct is 400VA An Antek AN-4235 would be exactly what you are asking for.

Tested secondaries parallel, but is secondaries series when I use it: Oops! Bit of a hard sell to say that doubling the length of the copper doesn't have any effect.

EDIT:
There was increased loss at secondaries, not at primaries. I've still got the Primaries parallel like the test conditions. So the discrepancy wasn't a factor of 2, but rather somewhat less. It is quite confusing to figure out specs that have a "somewhat" in them.

[Tony]

@danielwritesbac, testing and normal use are two different things.... there is really no comparison.....

[ncg2013]

I have a few questions....

In the build thread people are talking about a 2x45 Vac transformer 800va , if the speakers were 4 ohms what output power would that give? I wanted the amp to be flexible so I could use different speakers, maybe 8ohms at a later stage. I noted in one of your posts you said 45vac is too high for 4 ohm loads, what would happen? blow up?

2nd question (of many, yet to come). The PSU BOM states capacitors with a value of 10k uf to 22k uf, but some members are talking of just using 6k uf, what would the higher capacitance give, sound wise?

3rd question what is a snubber?

4th question - looked at some capacitors on Hi Fi collective and they we £15-£30, whilst others are talking of using 5usd capacitors. Would using more expensive be better, does it make any difference in a PSU? or should high quality caps be reserved for the amp?

That will do for now, hopefully be able to order something soon.

[ostripper]

Ahh , question 1 is very timely. I am about to build some "mini badgers" in bridge mode for my sub. The trick for 4R seems to be lower Rail voltages. At 40-0-40VDC the badgers SOA doubles and at 30VDC doubles again. 2 bridged Badgers running with 35v rails would most likely drive a 2-4R sub at over 300W !!
(BTW , this would equate to a 26-0-26Vac trafo).
The SOA of even the lowly onsemi NJW pair is close to 7 amps at 30V(DC).
Of course , using the more robust MJL's will increase thermal dissipation and these have even higher SOA.


Q2 - Smaller main supply caps = more ripple.For 8R loads , I use 10Kuf x2 per channel. For lower impedance's, 15-20Kuf is suggested.
6Kuf would turn this amp into an OEM (higher ripple - less "punch")

Q3 - The "snubber" Snubber - Wikipedia, the free encyclopedia
in this amp (the 2 diodes) keeps the inductive load of the loudspeaker from destroying the output stage. The 3 amp diodes should have a short reverse recovery time. You see this technique applied to relay coils as well.

Q4 - Some audiophiles swear by esoteric capacitors. My take is that low esr
and MTBF are the main considerations.
The input cap should be high quality polystyrene ,very stable over time. The miller caps in the VAS should be silver-mica - very stable (temperature and value) no parasitic resonances.
ALL power supply caps can be any properly derated (panasonic , nichicon ,rubycon), low esr electolytic.

OS

[solid]

OS,

Can you clarify a doubt...
In the BOM, R8 & R9 are specified as 100R 1%.
I'm wondering if you meant for R15 & R16 to be 1% rather than R8 & R9?