How to destroy your high-end audio equipment...

Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
Black Gate suppose to have a nice cap for decoupling, an N or Nx type. Well i tried four of them, leaving a big hole in my wallet for 4 caps.....

I soldered them near the THS4032 opamps in my modded Philips CD304mk2(clock PS diodes) When i started the player i thought something was wrong: no 3D sound anymore, flat and narrow (to the sides) it was allmost mono. I didn't want to burn them in, it was so terribly bad sounding. Replaced the caps with standard blue philips lytics, with a second smaller wima mks cap and a bead/1ohm resistor inbetween on rails.

Yes, now it sounded nice again!
 
I'm a little nuts, but there really was a change...

After looking at your chassis picture I might suggest that you also use cable ties on the caps to draw them closer together. This will have the effect of extending the low frequency response. White ties seem to make for better sound.

Oh burnedfingers, you are a scamp, aren't you?

To tubee - Glad to see you got things back to normal.

To Pete B. - I would LOVE to do the null test on my XL280(s) but I cannot figure out how to make an XL-10 switch box from Haflers schematic...
Luckily I have not had 1 problem with my 110, ever, until the tape monitor selector started distorting and switching out 1 of the channels randomly. I totally agree about the cheap RCA jacks, I would swap them out for good ones if I could find exact replacements, one is already broken off - but they work fine.

To those who want me to get this thing on a bench -
I deliver pizza. Tips today are less than they were 12 years ago, but gas is 3 times more. I have made less every year for the last 7 years, pizza men don't get raises. I think 'big oil' owes me a benchtest, but he doesn't return my calls. The only reason I have what I have is because I share a little apartment with my 2 brothers (I used to live alone).


People, I appreciate the ribbing that I'm getting more than you might think, maybe I should try to explain a little more.

I prepared for this project very carefully, I researched and chose the caps carefully, always mindful of the changes new caps had made to my power amp.
When I got her apart, I installed the Elnas and FM caps as planned, but as an afterthought, popped in the WIMAs because I had them sitting around from a TV project. They tested 10.3 and 10.5 nF.

When I powered up, the difference in sound was so horrendous that something in my head just could not believe those perfect little WIMAs could do such a thing, so in a panic I blamed the 'larger' change of the 12 power caps.

After my initial post, I realized the only way I could confess would be a bit of awkward stammering and looking at the floor.

As I said, I do not want to take part in ignorance, bias, or misinformation, hence my awkward confession.
----------

BUT,
All that being said, the more music I test on the system, the more I think the rest of those ICs are on their way back in.

I didn't think so at first, but just listened to a song that I am very familiar with and the top is still 'not as bright'.
Oh well, at least that will make the WIMA lovers out there happy.

The last time I inquired, the nearest shop that would benchtest my equipment was too far away for me to drive my beater.
 
Checking the bias requires only a multimeter. Adjusting it requires only a screwdriver. Nothing weird or complicated involved.
I pick up screwdrivers off the road where they've fallen off the back of trucks. I have yet to find a multimeter that way--I bought all mine. For that matter, I bought a lot of my screwdrivers, too. You can never have too many tools.

Grey
 
Idle current.

I once saw a post on DIYaudio where a guy was going berserk because many people referred to the idle current as the 'bias'.
Hilarious.
Do you guys love this nerd stuff as much as me or what? :)

At the moment, I am in safe mode concerning the IDLE current (Ok, I won't shout anymore), but I will crank it up a little when I know everything is A-Ok.

The Hafler XL-280 runs 3 pair of lateral Hitachi mosfets at roughly 300mA. Ideally, they share and each gets about 100mA putting them at zero tempco. Learned much of what I know from mlloyd1, and I have been reading a lot of Randy Slone.

XL280 #1:
My baby with NEC JFETs is set at 303mA/311mA per each side.
The DC offset is -3.5mV/-13.3mV

XL280 #2:
My other, eBay XL280 with Toshiba JFETs is set at 300mA/306mA and was reduced from its original 400mA+ settings.
The DC offset is 20.5mV/-20mV

I like my original XL280 better, and it does have new electros with the exception of the big 7800s. The sound is a bit more rich.

Still listening...
 
Imo low ESR only is not the secret of the modern Lytics. It is imo how the ESR curve is gradually expanding from the low audio frequencies. This should be rather smooth.

Further more every cap has its best field to use.

I modded a Quad pre and end amp. 303/33 Replaced all old 25 yo caps with Nichicons. It sounded better, also because bias was completely wrong (to low)

The owner is very glad with it again!
 
The last time I inquired, the nearest shop that would benchtest my equipment was too far away for me to drive my beater.

Not trying to pick on you here but....

So far according to prior posts you have massed an amount of different capacitors of which is more than what it would have taken in gas and time to drive to a shop and have the frequency response tested and plotted. Personally I do not buy the "I can hear the difference" claim.

I took the time and changed caps in the power supply of both a preamp and a power amplifier with several different brands and of different quality. All I can say is I must bow to your highly developed sense of hearing because I cannot tell any difference.

I do however think that maybe you could have a build up of wax in your ears which may be the reason you think the performance has been de-graded.

Hafler isn't considered by some to be a high end piece of gear because of its shody construction. I am of this belief.
 
Re: Idle current.

HaflerFreak said:


I once saw a post on DIYaudio where a guy was going berserk because many people referred to the idle current as the 'bias'.
Hilarious.
Do you guys love this nerd stuff as much as me or what? :)

At the moment, I am in safe mode concerning the IDLE current (Ok, I won't shout anymore)...



I think you will find that most texts refer to the process of pushing the control pin in such a direction (depending on whether the device is enhancement or depletion mode) that it conducts "biasing," not "idling" the circuit. The bias current is what you're measuring when you read the voltage across the Source/emitter/cathode resistor. 'Idle' refers to a circuit that is experiencing no signal; it is still biased, however, even at idle. The term Iq refers to the quiescent bias current--the property under consideration. It is measured at idle, but that's not the same thing as saying that you're measuring 'idle.'
If you can find a text--or for that matter, anyone who's been at this a while--that refers to bias as 'idle' then by all means be happy...just don't expect others to go along. If this qualifies as making me--or others--nerd(s), then that's okay...just think of what would happen if you walked into a hardware store and asked for a "tink." You wouldn't get very far until the salesman realized that you wanted to buy a hammer and that 'tink' was the sound the hammer made when it hit a nail.

Grey
 
it is still biased, however, even at idle

At least that much I understand.
Hello Grey, my comments weren't meant to offend in any way, I was just referring to the fact that on this amp you set the bias by measuring the idle current to the mosfets.
I am far too uneducated to know what I should be measuring across transistor legs.

As for the nerd comment, I thought it was good to be a nerd!

I used to be a fantastic nerd until depression took my life. I read day and night, got good grades, took all the honor classes. I loved, and still love, learning. You should have seen me go when I had a chance to go to college - straight A's.
I totally loved 'Revenge of the Nerds' back in the day.

to burnedfingers -
Yeah, the caps are a bit of a waste of money in my situation, but I get them all mail order. After 5 years of working with no vacation, I figure 15 or 20 bucks worth of caps provides me with hours of education and fun, so it's worth it (until I blow something up). I have had some amazing successes with a couple of high-quality Mitsubishi CRTs - finest picture in town.
I could probably hunt down a place to get my stuff checked, but I never now how much money I will have from day to day. Much of my spare cash gets tied up in covering for my deadbeat brothers.

You CAN hear the difference!

And no, there is no wax in my ears, I checked.

Put Panasonic FCs in for the 8 rail caps, you can definitely hear the difference.
The FCs are airyer, more spacious at the top, but with very anemic bass compared to the Elnas.
Overall, I like the FCs better so far, but they still aren't IC quality in their sound.

It's almost amazing the difference between the three caps I have tried, each has a different character.

- but the ICs are still on their way back eventually... :smash:
 
If you have access to a scope, you want to check for oscillation on the supply rails. I have a discrete buffer under evaluation powered by a discrete linear regulator. I measured its THD using the equipment at work and got very disappointing results. The culprit was oscillation on the supply rails caused by a big low ESR bypass cap. I isolated the regulator from the rail (and cap) using a couple of ohms and used a much smaller, higher ESR cap directly at the regulator output. Voila! The oscillation cleared up and so did the THD. Three terminal regulators can behave the same way.
 
muting fet in dh110

back when the dh110 was a current product I did warranty repairs on Hafler gear. and I actualy still use a dh110
if you are getting no output or strange sound, the muting
jfets are probably bad . this is a very common problem
I have changed them several times in my own unit
and i finaly just removed them.
btw the mod that made the biggest improvment in mine
was to change the output coupling caps to a good
polypropylene type .I did this by mounting them to a pc
board standing verticaly on the rail running down the center of the chassis. i also added another little board containing some larger power supply filter caps Ii think they were 15k uf.
 
110 mods

Hello,

To amp-guy: Now we're talking! I am very interested in any simple upgrades to this preamp.

Did you say you used 15,000uF caps in place of the main 1000uF caps? That seems a bit overkill - like you would need a soft-start for the preamp! I have considered going to 1,200uF, but after my latest recap, I think I might be happy enough to leave it alone.

Currently, everything is very well with me and my 110.
I have never had a single issue with the mute circuit, and I just replaced the bipolars in that circuit. The JFETs are original NEC K163's, I have some 2SK170's just in case....

I have not had 'weird' sound unless you consider the major differences I have heard between cap combos.
In another post, I described some of the caps I tried and the differences that I heard between them.

No one replied to my post when I decided that, hands down, the best sound I got from this preamp was with 'cheap' caps.
After trying various high-quality caps, in frustration I decided to simply match the specs of the original IC caps. The best matches for the 1000uFs were Nichicon VR (standard 85 deg) and the original IC RMR series matched up to Nichicon VZ's.

I then decided to follow the schematic and use 'low ESR' caps for the rails. The only caps I found listed as 'low ESR' and not 'low impedance' were cheapo XICON 25V 470uF from Mouser.

Not to make anyone cringe, but I tried XICON NPRL non-polars for the 22uF coupling caps. They claim to have 'excellent frequency characteristics in the audio range'.
They do.
I don't know if I have ever heard this pre sound better!
Especially in the bass, it is tight, fast, powerful - amazing.

I urge ANYONE with this model preamp to at least try the XICON NPRL's if they are going to change the coupling caps C22/122.

Merry Christmas everyone!
 
also, there are (believe it or not) counterfeit caps out there. there's a listing on www.badcaps.com. there's also a bit of a problem with a lot of companies using bad formulations for the electrolyte.
another thing to check (on caps that have them) is date codes. usually date codes are a 4 digit number (such as 9835, which would be the 35th week of 1998) or a letter digit digit, where the letter is a year, and the digits are a week. for the L-D-D code, you would have to look at a data sheet to find out what year the letter corresponds to. if a "new" cap is more than 3 years old, it probably needs reforming at it's rated voltage. lytics are an electrochemical device, and they have a shelf life of usually 5 years.
 
Member
Joined 2005
Paid Member
Haflerfreak - If you really want to get into your amp, do as GRollins says and crank the Bias. Crank it up till you can't touch the sinks for more than 10 seconds (that's around 60C) - that's what I would do.

Ahh, maybe not that high, but try it higher. You'll get that sparkle back.

I tend to think of an amp's "Bias" like a projectors' "Focus".

I have a ripped up Hafler 9130 that I am gutting. A friend accidentally spilled a diet coke into it and destroyed it. Planning on using the chassis for a nice UCD400. I still have not yet heard a functioning Hafler.

BTW - Lisle? Are you a Four-Laker? Lived there 9 years ago. Did the Volleyball Blind Draws, the Pub, Last Fling...I may move back someday soon.
 
Re: 110 mods

HaflerFreak said:


No one replied to my post when I decided that, hands down, the best sound I got from this preamp was with 'cheap' caps.



Many designers sit down with a piece of paper and design a circuit, then go to the bench and build it. If it works properly and tests well, they call it a day and (if they're a commercial outfit) start shipping product.
More careful designers sit and listen to the circuit, making changes according to what they hear. Whereas the bench designer is likely to be of the opinion that one 10000uF cap is as good as another, the more careful designer knows that parts choices make a difference. The process of finalizing parts choices according to the actual sound of the circuit is known as "voicing" the amp.
One of many possibilities is to use a "slow" sounding cap to balance a rather bright sounding circuit. This is often seen in high feedback designs, for instance. Although technically inferior, the slow cap actually works to the benefit of the circuit overall.
In your case, you may find counterintuitive combinations that work better than they should. It's just one of those things that constitutes art rather than science, since even careful mathematical analysis won't answer the "Which 10000uF cap should I use for this circuit?" question. The cap's specs can give you hints, but unfortunately they won't give you a complete answer...as you're discovering. Don't fight it. It's just the way things are. Granted, it takes time, effort, and the outlay of money, but the results can be quite gratifying if you can get things properly settled in.

Grey
 
Hey guys,
You guys are great, I never get enough of this stuff – even when I don’t really get it.

To unclejed613:
I know date codes, all of the caps I have been using are late ’06 to mid ’07.
I doubt the caps are counterfeit, I have little choice but to order everything from Digi-key and Mouser.
- I love that Pac-Man quote.

To john65b:
I think I have sort of wanted to hear that suggestion concerning bias. I have had serious second thoughts about touching the bias. Without a sig gen / oscope or distortion analyzer, I figured I may never get my sparkle back. Both of my XL280’s sound a little dull since I set the bias to 300mA – but they run very cool. I have considered trying about 330 to 350mA. I am quite concerned about shortening the lives of my matched MOSFETs.
FOUR LAKES:
My mother lived there, I delivered pizzas there until recently.
You should see it! THEY DESTROYED IT! I and my brothers call it ‘The invasion of the body snatchers’ – every land owner in DuPage county is clearing any sign of nature from their properties. It’s disgusting, they cut down 300 year old oak trees to plant sod and mismatched annuals. I HATE living here (and I grew up in Downers). It’s very sad to see it all go, the nature, the beauty, the open space.
It’s just traffic and big ugly houses now.

To Grollins:
Grey, you are truly awesome. Without a formal education in this stuff, I had to assume many of the points that you made in your comment. As others, and datasheets, have pointed out, there is a radical difference between caps. I have always known I was not imagining many of the ‘changes’ that I have heard.
The fact is I still don’t think I’m happy. I would really like better quality caps in my preamp than XICONS, but I have not found out how to get small quantities of Nichicon Muse – the ones I would like to try next.

About the ‘slow’ cap: I think this is what I am finding in my XL280! I was very excited to read this, because in my stock XL-280 I found one driver card had 105 degree caps and the other had 85 degree caps. I know 105/85 degree caps have different specs and the 105 would normally be ‘slower’. With new, matched caps, my XL-280 does NOT have the deep bass that it once did – except when I had XICON high-temps in it. The caps in it right now are UCC KZE low impedance caps. It sounds good but a little thin.

The cap experiments are NOT over. I remember exactly how good this system sounded over the years, people used to tell their friends to come over and hear it. It was like I was running a museum exhibit for awhile. I had people tell me they had never heard sound like this system once made.
I will find it again!
 

Attachments

  • caps.jpg
    caps.jpg
    51.6 KB · Views: 855
Member
Joined 2005
Paid Member
Too bad about four lakes...I was looking forward to eventually playing V-ball there again someday.

What Maggies do you have? I have MGIIIa and love them, I had MMG and still have a pair of MMG-W (I need to sell these to thin the herd)

As far as sucking life out of your MOSFETs, I wouldn't worry much until you get quite hot. Crank Bias up until your heatsinks get to 110 - 120 F and you should not be sucking much if any life out of your output devices....of course YMMV...
 
Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.