Deleted
Deleted Member
Posts: 0
|
Post by Deleted on May 1, 2008 22:09:39 GMT
Holland
The original circuit was shown in a graph by Silicon Chip as having a -3dB point of > 100KHZ into 8 ohm headphones, and approx 170KHZ into 32 ohms. 600 ohm headphones figure is >200KHZ. Bear in mind that this is with a RLC Zobel network, so Mike's current version would be even better. I use the Toshiba 2SA1930 and 2SC 5171, which have an Ft of 200MHZ, and better HFE than typical BD139 and BD140, which I have read vary greatly as regards Ft , between manufacturers. In fact, it is hard to find data regarding the Ft of the BD139 and BD140.
If you are correct , about the device being AD8067, then it isn't really suitable in this present circuit.
Alex
P.S. You are pretty astute for a 108 year old person ! ;D
I can email you some of the original article if you are interested.
|
|
|
|
Post by PinkFloyd on May 1, 2008 22:31:39 GMT
If I read the package markings correctly (picture is blurry), it's the AD8067. It's not stable at gains below 8 unless compensated. That rings a bell..... yup.... that's the one AD8067.... not sure what the stock gain of the Jaycar is but I'll be increasing it to around 6 or 7 tomorrow.... just feel I need a bit more volume sooner on the dial. |
|
|
|
Post by PinkFloyd on May 1, 2008 22:34:33 GMT
By the way Alex... in it's current guise, with AD744 biased into class A, and even without the much praised JLH this is one CORKER of an amp and would give most commercial amps one helluva good run for their money and on a bang for the buck comparo beat all of them hands down, no contest. Man, it really IS good being alive again I'm sooooo much better with that crap out of my system.... positively raring to go!!  |
|
|
|
Post by holland on May 1, 2008 23:01:58 GMT
Thanks sandyk. I didn't want to enter my real DOB, though it probably doesn't matter much.
Yes, please e-mail me the article or parts of it.
This sounds like a great bargain. I'm waiting for it to become available again from Jaycar, I saw the other link, but the prices seemed to be double that of Jaycar. I don't think I can source the parts and etch my own PCB for less than the kit from Jaycar. Perfboard by itself would run me almost $5-6 USD.
Can someone try the 2SC3422 and the 2SA1359 as the output transistors?
|
|
leo
Been here a while! 
Team wtf is it?
Posts: 3,638
|
Post by leo on May 1, 2008 23:32:53 GMT
AD744, is this the one you bypassed the internal output section and run the output from the internal input stages? If that makes sense  I reckon that would work well in a amp like this |
|
leo
Been here a while!
Team wtf is it?
Posts: 3,638
|
Post by leo on May 1, 2008 23:43:12 GMT
BTW its excellent to hear your getting back into things Mike |
|
Deleted
Deleted Member
Posts: 0
|
Post by Deleted on May 2, 2008 0:42:03 GMT
Tweaking the SC HA- Silicon Chip This may be easier to read than the supplied photocopy?  |
|
|
|
Post by PinkFloyd on May 2, 2008 3:07:40 GMT
AD744, is this the one you bypassed the internal output section and run the output from the internal input stages? If that makes sense I reckon that would work well in a amp like this That's the one.... the bend the leg over jobby  and yes, Leo, it works extremely well Will fire in the LM4562 again tomorrow as it may be a lot better now I've got rid of the suntans and the few other bits and bobs..... flippin' great amp to tweak and good fun too Night |
|
|
|
Post by canjunkie on May 2, 2008 12:23:34 GMT
Oh dear........ Jaycar have refunded my CC for the cost of the amp & psu modules.  If I can re-order at a later date, incurring another delivery charge isn't too bad when the units are so cheap. Alex, do you have any contacts that can tell you whether they will be re-stocking and if so whether their initial etd of June has slipped? |
|
|
|
Post by dc on May 2, 2008 12:43:24 GMT
looks like i've missed a lot in 3 days Mike, it's funny, but the AD744 class-A and the AD843 and two of my favourite opamps the equal third being the LME49710 which is a cracker of an opamp too when a cooler presentation is needed (if i'm not mistaken it's just the single version of the LM4652) nice to see so much extreme tweaking done on this basic and inexpensive aussie product |
|
Deleted
Deleted Member
Posts: 0
|
Post by Deleted on May 2, 2008 21:26:26 GMT
canjunkie
Holland mentioned that the similar kits from Altronics were very expensive. This surprises me as their Aussie prices are very competitive with those of Jaycar. Their HA kit is on special this month, and even cheaper. Perhaps it's a higher P and P charge ? Holland may care to elaborate.
I am on good terms with the local nearby Jaycar managers, and next visit there I will try to find out more.
Alex
|
|
|
|
Post by canjunkie on May 2, 2008 22:47:25 GMT
Alex, I've emailed Altronics with a stock enquiry and request for shipping costs. 25AUD$ works out to be about GBP11.50 which is a little more than the approx 9GBP Jaycar price. Still looks an absolute bargain. If the news isn't good from Jaycar and Altronics do have stock I'll order Amp & PSU from them.  Thanks for all your help Alex. I'll email a modified JLH PCB which follows your suggestions for modding when I get it finished later this weekend.  I wouldn't want to offer it to anyone else before someone's who really knows what they're doing has looked it over . Canjunkie |
|
|
|
Post by holland on May 2, 2008 23:07:54 GMT
Oh, it's not horribly expensive, but it is relatively more. For me, shipping to USA from Jaycar.com, the kit is about $17.50 USD with about $7.50 USD shipping. The PSU is another $10.50 USD with the same amount of shipping.
Altronics was about $25.00 AUD (after special) and the PSU is about $17 AUD, I don't know what shipping is as I would have to contact them to order. The USD and the AUD are very close now.
The Jaycar kit, I don't think I can source the parts for less than the kit (and etch a PCB). The Altronics kit, assuming shipping is $7.50 as well, I can probably source and wire on a breadboard or perfboard for about the price that it sells for.
While I'm sure it's a nice unit, I'll wait for it to come back in stock @ Jaycar. If I'm desparate, I think it would be cheaper for me to source the parts (I already have 95% of it in my parts bin or 100% if I skip the zobel) and wire it on breadboard or etch a PCB. I'm a fan of the zobel, so I won't skip that. I've seen too many projects go into oscillation from inductive or capacitive loads.
|
|
Deleted
Deleted Member
Posts: 0
|
Post by Deleted on May 2, 2008 23:14:09 GMT
canjunkie and holland
I have sent you an email.
BTW, in the attached photo I have used 2SA1930 and 2SC 5171 for the larger devices. I understand that Allan uses BD139 and BD140, which aren't such a tight fit. The 2SA1930 and 2SC5171 need to be rotated, if replacing BD139 and BD140. Please check their data sheets.The supply input end has been altered by Allan, to accomodate a 3pin terminal block with 5.08mm pin spacing
Alex
|
|
|
|
Post by holland on May 2, 2008 23:18:09 GMT
Been meaning to ask. I am a bit concerned with DC offset and drift. There's no servo or trimpot for offset adjustment. That might be something to take into account as one swaps opamps around. BJT input opamps (like the LM4562) might likely yield an offset from the higher Ib causing some slight oscillations. From a quick glance at the circuit, it doesn't look like legs of the opamps have balanced resistances. I've seen the LM4562 cause some issues in the past, in CMoy type configurations where it's most obvious that something is wrong.
|
|
|
|
Post by holland on May 2, 2008 23:20:36 GMT
I have sent you an email. Got it. Thanks much! |
|
Deleted
Deleted Member
Posts: 0
|
Post by Deleted on May 2, 2008 23:34:00 GMT
Holland
I will recheck the offset on my unit later today. You will have noticed that I have left the input capacitor in circuit. I have only tried OPA2134, AD8066 (using +-12V) and LM4562. The AD8066 was only tried with the original circuit using the RLC Zobel network.
There is an option shown for R3 and R6 on the previously posted blue "Tweaking" page.
Alex
|
|
Deleted
Deleted Member
Posts: 0
|
Post by Deleted on May 3, 2008 0:01:27 GMT
Holland
I have just checked the DC offset at the headphone jacks of the fully modified unit which may be going back to the U.K. shortly. There is 0.8mV on one channel, and 0.4mV on the other channel. The readings seem fairly stable. I don't expect a huge change with temperature. I am using 2K0 resistors for R3 and R6.
Alex
|
|
|
|
Post by holland on May 3, 2008 0:09:37 GMT
Thanks, Alex. I noticed the cap. I read the R6 addition as well. That blocks DC from the source, but it won't handle DC generated by the input bias current of an opamp. With a path to ground on the input leg, and an input bias current flowing, it's going to generate a DC voltage at the input of the opamp. With a relatively small gain of 2, it may not be that big a deal, but if gain is increased it may be an issue.
It might be able to be reduced by increasing the 100ohm resistor and lowering the 47pF cap. However, there's a tradeoff for resistor "noise".
The other concern is the thermal drift of the transistors over time. That may not be an issue since the voltage gain stage isn't discrete, but I'm not sure.
Some of the faster opamps may be applied as well if a multi-loop is hacked into the circuit, to reduce the bandwidth of the fast opamps (LM6172, AD8066, etc.). I probably wouldn't use anything over 50MHz without a multi-loop. Just a thought.
|
|
|
|
Post by holland on May 3, 2008 0:10:21 GMT
There is 0.8mV on one channel, and 0.4mV on the other channel. The readings seem fairly stable. Thanks, that looks quite good. |
|
Deleted
Deleted Member
Posts: 0
|
Post by Deleted on May 3, 2008 1:10:27 GMT
holland I use the 2SA1930 and 2SC5171 200MHZ Ft transistors in my version of this HA, as well as my preamp. An earlier version was used with an AD8066 as well as those O/P devices, with excellent results. From memory, I think I put another couple of 100nF poly capacitors under the PCB for AD8066 stability. See attached. Alex  |
|
|
|
Post by holland on May 3, 2008 1:59:59 GMT
Yes, you're right. I had my transistors of choice in mind which have an fT of 100MHz. 1st stage faster than second is bad. I already have mods planned out before even buying the kit  |
|
|
|
Post by PinkFloyd on May 3, 2008 21:18:07 GMT
Mike, it's funny, but the AD744 class-A and the AD843 and two of my favourite opamps Always been two favourites of mine they just "sound" so right don't they? The best I heard was the AD-811 / 744 hybrid, IIRC it was Walt Jung who came up with this....... the bugger is I can't remember how to hook the two of them up and can't find a diagram anywhere..... I'm sure Leo or Rick may have it kicking about in their brains, I really would like to put those two together again (744/811) but need the hook up details. Maybe I'm dreaming? Seems like an age ago but I'm sure I tried a 811/744 hybrid in the WNA and nearly creamed my pants. Anyone who can jog my old memory re: 811/744 hybrid will get a free sausage roll and a cup of tea |
|
leo
Been here a while!
Team wtf is it?
Posts: 3,638
|
Post by leo on May 3, 2008 21:41:29 GMT
I think you mean running signal out from pin5 of the AD744 (before its internal output stage) and running it into the AD811 (pin3) so using it as a current buffer?
Be fiddly to do with these op-amps being singles and the Jaycar using a dual, its do able of course if you have the patience
|
|
|
|
Post by PinkFloyd on May 3, 2008 22:04:04 GMT
I think you mean running signal out from pin5 of the AD744 (before its internal output stage) and running it into the AD811 (pin3) so using it as a current buffer? Be fiddly to do with these op-amps being singles and the Jaycar using a dual, its do able of course if you have the patience Yes! More here: www.audioxpress.com/reviews/media/AE200GG2.pdfI do have the patience and well worth a try I think Leo. |
|
and yes, Leo, it works extremely well 
I wouldn't want to offer it to anyone else before someone's who really knows what they're doing has looked it over 