leo
Been here a while! 
Team wtf is it?
Posts: 3,638
|
Post by leo on Feb 28, 2011 21:57:14 GMT
Ok ta, the Minimelf suggestion was just something for those not willing to spend more than the cheapo ebay offerings , going from past experience smd resistor type can have quite a noticeable influence on SQ, some smd resistors really do sound quite dire.
So I presume the mini V would be the better choice over the Dact types if wanting the best?
|
|
Deleted
Deleted Member
Posts: 0
|
Post by Deleted on Feb 28, 2011 22:10:06 GMT
Hi Chris. I think on this, it's different requirements that causing the confusion. As I'm using relays to switch the signal and ground, I only need the one pole, and the 5 positions to meet my requirements. The other pole is unused (though I will most likely double up) This switch could cope with signal (and associated GND) to 5 outputs , or 5 inputs to signal. The led side of it would need another pole. Aha, yes. You're doing it properly and I'm going second best, I forgot about your relays  For my bodgey method the following might do the trick, two of them coupled front to back... uk.farnell.com/lorlin/ck3610/switch-2pole-6pos-front-rear-drive/dp/1774550 |
|
Will
Been here a while!
Ribena abuser!
Member since 2008
Posts: 2,164
|
Post by Will on Feb 28, 2011 22:20:02 GMT
No no, not second best at all Chris  It's just two different approaches to achieving the same end result. I remember the question (relay switching vs. rotary switch) cropping up a few years back, and I think it was Rick who said he used to use relay, but went to rotary switch for simplicity. The aspect of having a front panel mounted selector switch, and having signal wires from the back to the front, and then back to the amp pcb can be addressed by using diy/Shaun style brackets, and an extension rod, if it's considered a problem. That switch you linked to looks spot on, and I reckon you'd be able to fashion a connector rod easy. |
|
Deleted
Deleted Member
Posts: 0
|
Post by Deleted on Feb 28, 2011 22:27:31 GMT
So I presume the mini V would be the better choice over the Dact types if wanting the best? Leo It would appear so on what we can presently see, however I doubt that DACT would be standing still in that area. So I would expect that that their current versions are likely to have.1% thin film resistors too.We need to remember the point of diminishing returns here too, as even the Valab attenuators are clearly superior to the blue Alps, and Shaun has reported that the DACT2 is only a little better than the Valab. Alex |
|
Deleted
Deleted Member
Posts: 0
|
Post by Deleted on Feb 28, 2011 22:32:44 GMT
Will,
You read my mind! LED switcher at the front with the input switcher to the rear via a long coupler.
I'm also glad to hear that quality may not suffer after all.
Cheers,
Chris
|
|
leo
Been here a while!
Team wtf is it?
Posts: 3,638
|
Post by leo on Feb 28, 2011 22:41:11 GMT
Hi Alex, I just thought its worth knowing which is the best choice if wanting to go for the higher priced ones. We both know that in audio even the small differences can be a big deal to some people  Regards, Leo |
|
Deleted
Deleted Member
Posts: 0
|
Post by Deleted on Feb 28, 2011 23:05:03 GMT
Will, You read my mind! LED switcher at the front with the input switcher to the rear via a long coupler. I'm also glad to hear that quality may not suffer after all. Cheers, Chris Chris and Will RCS Radio came up with a simple PCB for the Studio Series Preamplifier, which operates the appropiate Input relay, as well as having front panel LEDs. It uses an onboard IDP socket and an IDP cable to the relay switching PCB. Unfortunately, I have lost the diagram that came with them,but someone should be able to work it out. Alex Attachments:
|
|
Deleted
Deleted Member
Posts: 0
|
Post by Deleted on Mar 1, 2011 0:00:51 GMT
Hi Leo how's it going? as far as i can tell the Mini V and DACT are both series as is the Vlab unless i've missed the shunt R on the Vlab. i may be wrong so feel free to correct me if I've blundered.  yes interesting as the Mini V and Dact both use the same Elma switch but from what i have read have tensioned the springs slightly differently. so any sonic difference would be down to the quality of resistors used and yes i do think that's going to make some difference (did not really believe that resistors made much impact on the sound until i tried some tants on the cathodes of some of my tube amp and heard the noise floor drop) also the construction quality of the boards may impact on the sound i guess to a lesser extent. I've not heard the mini V so can't say how it sounds but that could be interesting to compare at some point. I've read that the goldpoint is pretty darn good though but as Alex said it's a game diminishing returns.  I've schmooched round the net and found some comparisons. some like one and some the other (even read one who preferred the DACT because it was queter to use and did not annoy his wife(spring tension)) so if the mini v uses better resistors then in theory it should sound better.  i fear that my current set up has not the resolving power to clearly show the difference between Vlab and DACT. you may have noticed that i am planning (well mostly groping around in the dark) a PC hi res set up with hopefully a Buffalo at some stage to bump up the SQ maybe. i did opt for the DACT as long term thing in a kind of fit and forget way. it'll be interesting to try the Vlab again once I've sorted my front end just to see if there is more of a marked difference compared to the DACT or not. but in my present set up there was no huge jump in SQ from one to the other and the most obvious thing was the improvement to the top end. i did try an Alps Blue which sounded shut in in comparison to the Vlab IMHO. lets be honest if i could by something with the quality of the Vlab ten years ago for that price I'd be pretty happy (very very happy in fact).  so that's moved on quite a bit in a short space of time. it may be an interesting experiment to shunt the Vlab with a nice resistor to see what that does. shunts of course have there problems but I've had pretty good results from shunting steppers with a nice (box vishay) resistor. then maybe try the Alps shunted with the same resistor just for fun. TBH i just don't like what shunts do to the input impedance and SQ. i have used them but really hard not to as they can sound a bit unnatural to my not so good ears. i think that if i was going for the best VFM I'd opt for a six position switch (Elma) with maxed out high quality resistors (just for the HA not the PRE) as i don't seem to use much more than that on the switch at the moment. any way blah blah blah sorry but I'm interested in your thoughts switches don't even get me started on that  take care |
|
XTRProf
Fully Modded
Pssst ! Got any spare capacitors ?
Posts: 5,689
|
Post by XTRProf on Mar 1, 2011 1:16:59 GMT
Chong, for me that would definitely be a step too far£££££££££££! Oh, just to keep the dream alive as I'm not sure what's your objective going into this AHA. |
|
XTRProf
Fully Modded
Pssst ! Got any spare capacitors ?
Posts: 5,689
|
Post by XTRProf on Mar 1, 2011 1:18:17 GMT
Any Mk 3 AHA version with a thinner PCB, bigger pads and the offset adjustment for the op amp when using either AD744 or OPA134 to get close to 0 mV dc offset?  If you build it correctly using the LS devices you will not need any adjustment when using an OPA134 as recommended. If you can't get close to 0mV at a normal listening level, then you have something wrong !  Even the bipolar input LME49710 which are not recommended in this application were able to do 7 mV. The OPA134 are FET input with very low DC out. (check the data sheets.) Yes, I understand but just in case. Well, anything can happen, right? |
|
Deleted
Deleted Member
Posts: 0
|
Post by Deleted on Mar 1, 2011 1:26:11 GMT
If you build it correctly using the LS devices you will not need any adjustment when using an OPA134 as recommended. If you can't get close to 0mV at a normal listening level, then you have something wrong ! Even the bipolar input LME49710 which are not recommended in this application were able to do 7 mV. The OPA134 are FET input with very low DC out. (check the data sheets.) Yes, I understand but just in case. Well, anything can happen, right? Chong I will again state that if the Offset corrector when using a suitable device such as the OPA134, has excessive offset, then you have a fault condition that should be investigated ! One such cause has been found to be due to inadvertently swapping one of the 100R in the front end with a 150R during construction. A friend of mine who is very experienced, accidentally did that recently in a similar design. Alex |
|
XTRProf
Fully Modded
Pssst ! Got any spare capacitors ?
Posts: 5,689
|
Post by XTRProf on Mar 1, 2011 1:26:11 GMT
If you build it correctly using the LS devices you will not need any adjustment when using an OPA134 as recommended. If you can't get close to 0mV at a normal listening level, then you have something wrong ! Even the bipolar input LME49710 which are not recommended in this application were able to do 7 mV. The OPA134 are FET input with very low DC out. (check the data sheets.) Hi Chong what's the point in complicating things on the board further for very little gain? makes no sense to me if it ain't broke don't fix it. take care Oh, if there is the possibility of a MK 3 version of AHA, why not? I agreed with you when it ain't broke, don't fix it. But if we are not lucky we still have some dc offset. So it's still "broke". That's what improvement is all about when we reach the highend stake. I still can live with some dc offset but if there is a possibility of a Mk 3, we can improve the design further. Why go for the same for Mk 3, right? |
|
Deleted
Deleted Member
Posts: 0
|
Post by Deleted on Mar 1, 2011 1:36:54 GMT
Hi Chong what's the point in complicating things on the board further for very little gain? makes no sense to me if it ain't broke don't fix it. take care Oh, if there is the possibility of a MK 3 version of AHA, why not? I agreed with you when it ain't broke, don't fix it. But if we are not lucky we still have some dc offset. So it's still "broke". That's what improvement is all about when we reach the higend stake. I still can live with some dc offset but if there is a possibility of a Mk 3, we can improve the design further. Why go for the same for Mk 3, right? Hi Chong fair point take care |
|
XTRProf
Fully Modded
Pssst ! Got any spare capacitors ?
Posts: 5,689
|
Post by XTRProf on Mar 1, 2011 1:43:37 GMT
Yes, I understand but just in case. Well, anything can happen, right? Chong I will again state that if the Offset corrector when using a suitable device such as the OPA134, has excessive offset, then you have a fault condition that should be investigated ! One such cause has been found to be due to inadvertently swapping one of the 100R in the front end with a 150R during construction. A friend of mine who is very experienced, accidentally did that recently in a similar design. Alex Noted again! Thanks Alex. |
|
XTRProf
Fully Modded
Pssst ! Got any spare capacitors ?
Posts: 5,689
|
Post by XTRProf on Mar 1, 2011 4:47:03 GMT
Actually, I'm hesitating connecting the signal ground to the chassis ground. That's why I didn't put a connector at the GND position. Any advantage? I thought signal ground separated from the chassis ground is the best?
|
|
Deleted
Deleted Member
Posts: 0
|
Post by Deleted on Mar 1, 2011 4:54:45 GMT
Actually, I'm hesitating connecting the signal ground to the chassis ground. That's why I didn't put a connector at the GND position. Any advantage? I thought signal ground separated from the chassis ground is the best? Chong This is a proven design as verified by the several members who have constructed it. However, if you want to do things a lot differently, feel free to do so. Alex |
|
Will
Been here a while!
Ribena abuser!
Member since 2008
Posts: 2,164
|
Post by Will on Mar 1, 2011 7:08:53 GMT
Actually, I'm hesitating connecting the signal ground to the chassis ground. That's why I didn't put a connector at the GND position. Any advantage? I thought signal ground separated from the chassis ground is the best? If you don't fit something in the link, then you will have very high dc offset. Start with a link, get it working, and then tweak. As for v3 of the HA, it's not happening. There may be a slightly tweaked second GB for the PCB, if there is enough interest, but there is simply not enough space for offset nulling on the existing pcb, without a major re-design. If you wish to do it, get your pcb working, see what offset you achieve, and then fit a pot across the appropriate pins and dial in for 0v. Details of the pins in the appropriate datasheet. |
|
Deleted
Deleted Member
Posts: 0
|
Post by Deleted on Mar 1, 2011 7:31:06 GMT
Aside from the DC offset, personally, I wouldn't fit anything else then a link (or 0 Ohm resistor) , as Alex mentiones in his schematics, in that position.
I would not consider placing any resistance there.
Chong's idea of 'groundloop breaking' (where ground lift is usually made for) I would do between the ground connection of the PCB and the chassis in this case.
(based on the schematics that have been posted on here where the PCB is based on)
|
|
XTRProf
Fully Modded
Pssst ! Got any spare capacitors ?
Posts: 5,689
|
Post by XTRProf on Mar 1, 2011 8:53:28 GMT
Aside from the DC offset, personally, I wouldn't fit anything else then a link (or 0 Ohm resistor) , as Alex mentiones in his schematics, in that position. I would not consider placing any resistance there. Chong's idea of 'groundloop breaking' (where ground lift is usually made for) I would do between the ground connection of the PCB and the chassis in this case. (based on the schematics that have been posted on here where the PCB is based on) So as not to get confusion here, please note that I'm talking of the link between Signal Ground to Chassis Ground ie GND CONNECTOR, which is missing in my PCBA, to BOX CHASSIS. Of course, I will put the other "link" as what I can see everybody was referring to, otherwise there wouldn't be any sound coming from the HP due to open circuit if we follow the traces on the AHA PCB. I will put that link with OCC copper tinned with solder to prevent copper oxidation. Frans is talking the same lingo as me there. |
|
Deleted
Deleted Member
Posts: 0
|
Post by Deleted on Mar 1, 2011 8:54:20 GMT
Aside from the DC offset, personally, I wouldn't fit anything else then a link (or 0 Ohm resistor) , as Alex mentiones in his schematics, in that position. I would not consider placing any resistance there. Chong's idea of 'groundloop breaking' (where ground lift is usually made for) I would do between the ground connection of the PCB and the chassis in this case. (based on the schematics that have been posted on here where the PCB is based on) Frans Feel free to add to , or correct the attached. Alex In addition to what you have posted, my understanding of "groundloop breaking", is that it is there to prevent hum due to multiple earth paths. Although the centre tap of the transformers may be connected to chassis, there is no connection to mains earth , other than may come from the appliance it takes it's input from. In the case of modern CD/DVD players, there will be NO direct connection to mains earth when used with this HA. |
|
Deleted
Deleted Member
Posts: 0
|
Post by Deleted on Mar 1, 2011 11:32:04 GMT
What I found in the schematics is the following 'error'.
The HP return (L+R) is going to the input ground of the LTP.
So if you for instance mount a 1k resistor the return current of both HP channels will lift the voltage of the common from the LTP and thus the whole input is no longer referenced to ground but dependent of the OP voltage.
This is unwanted to say the least.
So 0 Ohm should be mounted.
a possible groundloop in this amp could be coming from the leakage current of the transformer(s).
If the transformer leaks this small current will enter the common of the amp and this leakage current travels through the chassis if connected to safety earth and also via the line input's ground connection (shield RCA plug).
depending if the casing of the amp is connected to safety earth, which might be broken by external power supplies where the safety ground is not connected, currents will be diverted to safety ground.
In this case the leakage current will pass through the shield of the audiowiring (RCA) and induce low level hum which MIGHT become audible.
breaking of ground loops should be done between RCA input shield and common, so the leakage current will travel through safety ground and not though shield.
highly depends on the source and how THIS is connected to safetyground..
|
|
jonclancy
Been here a while!
Mr. Ripple Eater
Amateur EAGLEist
Posts: 1,131
|
Post by jonclancy on Mar 1, 2011 12:07:56 GMT
I've stuffed resistors in mine so far, doing a bit at a time. Having a class A already built up does take the urgency off for me How's everyone else doing? Hi Will, All, An update from me: All resistors and trimpots stuffed. And, thanks to Will lending me his DCA (cheers!) and me seeing just how good it is, I decided to invest in the discounted set of two. So, I can now match my caps as well (yes, I know, I know!!). This will be done today and they will be stuffed in accordingly. After that, it's on to matching the transistors. Pot and trafo-wise, I am still in the deciding stage. However, I am looking at Nuvotem Encapsulated items from RS, as I have used them before and they are top quality. Either 2x18 or 2x25 / 15 or 30VA uk.rs-online.com/web/search/searchBrowseAction.html?method=getProduct&R=2238493uk.rs-online.com/web/search/searchBrowseAction.html?method=getProduct&R=2238550uk.rs-online.com/web/search/searchBrowseAction.html?method=getProduct&R=2238500uk.rs-online.com/web/search/searchBrowseAction.html?method=getProduct&R=2238566I have the case already. I wil be looking to use this as a preamp, too. Either single or dual inputs, but not too much more complex than that (unless I go down Will's route!! ). Cheers Jon |
|
Deleted
Deleted Member
Posts: 0
|
Post by Deleted on Mar 1, 2011 12:30:18 GMT
Chris and Will RCS Radio came up with a simple PCB for the Studio Series Preamplifier, which operates the appropiate Input relay, as well as having front panel LEDs. It uses an onboard IDP socket and an IDP cable to the relay switching PCB. Unfortunately, I have lost the diagram that came with them,but someone should be able to work it out. Alex Anything to do with this reference Alex? sc 01109053 $ 8.50 3039B 6 LED, 6pos Rotary Switch 2005Oct I'm assuming it refers to a Silicon Chip article? |
|
jonclancy
Been here a while!
Mr. Ripple Eater
Amateur EAGLEist
Posts: 1,131
|
Post by jonclancy on Mar 1, 2011 13:20:56 GMT
Just checked my Tubeshunter-supplied Toshes.
5171s: 179, 179, 158, 149
1930s: 156, 155, 148, 147
About to check my Digikey supplied pieces. I'm using a permanent silver Sharpie marker to record the values on each piece for future reference.
Seeing as Tubeshunters prices are similar to Cricklewood, he's a good option if you don't want to buy a job lot and match yourself.
Cheers
Jon
PS, I can confirm that two DCA55 from different batches measure within 1 Hfe of each other. Highly recommended bit of kit!
EDIT: My Digikey 1930s were all around the mid 140s or so. My 5171s were much higher Hfe - around 190s to 205. I take it we want the 1930s and 5171 to at least be in the same ballpark?
|
|
jonclancy
Been here a while!
Mr. Ripple Eater
Amateur EAGLEist
Posts: 1,131
|
Post by jonclancy on Mar 1, 2011 14:07:08 GMT
Just set up a small rig to measure the 100uF 'lytics (Panny FM). There is a small range in capacitance, as you'd expect. The rig allowed me to plug in caps and test as fast as I could replace them and mark the value on the side. I tested over 30 caps in 15 mins. I intend to mirror values on the board and between boards, assuming I have 4 of each value. Of course, this might be overkill, given that my source, cans and lug'oles aren't matched anywhere near the accuracy I have put into this build. But it didn't take much time at all, and will probably supply you all with some ribbing material!!  |
|
It's just two different approaches to achieving the same end result.
so that's moved on quite a bit in a short space of time.
