DIYHiFi.org

1.
I was too gullible. With the right output transformer, it did at 1200 MHz.
But I got it stable.

2.
In 2001, I wrote two articles on crystal oscillator design for DUBUS.
DUBUS is a ham radio periodical.
Yesterday, I converted the first of these to a .pdf. It can be found at

http://www.hoffmann-hochfrequenz.de/dow ... _Teil1.pdf

3.
Since Jocko came 2 weeks too late, I had to d.i. myself and converted
Jung's positive and negative regulators and his low noise reference to
stamp sized boardlets. Was the minimum effort.
This evening, I populated the reference and the positive regulator
and they seem to work as described. Performance characterisation
will have to wait a week or so.

Used stinkin' SMDs, of course.
I don't plan to sell boards, but will put the photomask as pdf on
my web site. After all, this is diy. It's a 160*100 mm Eurocard
with 2 or 3 samples of each. It is 2-sided, but the bottom side
is unetched ground. The I/O pads are on 100 mil centers so the
stamps fit like ICs on predrilled board etc.

4. For those in search of a project: A DIRTY power supply would be great.
It would add some noise/sine from a function generator etc to its
dc output voltage. Then one could easily sweep line suppression.
Looks like a job for one of these chip amps.

regards, Gerhard

_________________
Everything has been said already - but not yet by everyone. (Karl Valentin)

SMD for RF........yeah. But for "audio", I'm off of them. Most of the problems are self-heating of tiny parts, and excess noise of those crappy thick-film resistors. I tried making a SMT version of some of my old designs, and they were all noisier. Not a lot, but enough to annoy me.

(I get annoyed easily. Who could imagine?)

I may eventually try my crappy design for a negative regulator. However, it was only intended to supply +5 V from + 9 V.

Jocko

Jocko
I found the same. Some time ago I started a thread asking about smt resistors. Dropped it. Even with a low Tc R, the self heating changed the R value more than I liked. I did find non-magnetic resistors, didn't buy any.

In the line stage these were to be used in; it's designed to get the thermal tracking of the Rs as close as possible. This is almost impossible with smt unless the current through them is dropped to the point that the circuit won't work the way I want it to.

I do use smt semi's, heatsinking these can be a challenge.

Long live through hole!

We wouldn't use thick-film through-hole resistors, so why use a teeny-tiny version that gets hot?

I remember when what used to be called Roderstein dropped their MK3, and said the MK2 would work just as well. I complained to one of their engineers, and he failed to comprehend why I was upset. (Not that it would have changed the demise of 1/2 W resistors.) I bought as many I could get for the stuff we were producing. I managed to add the left-overs to my personal stash. And, no.........I am not sharing.

However, the latest PCBs I have made use RN55 size. Anyone want them with RN60 size? (Too late for this regulator, and the clock will have to use SMT, only because of size. Excess noise be damned.)

Jocko

Thank you Gerhard, I've been looking for it for so long time!!
...and thanks also for the other infos ..really useful!


When dissipation is low and smt is needed (read: RF signal circuits) I use these:
http://suppliers.digikey.com/Manufactur ... US&lang=en
http://www.susumu.co.jp/english/product/index.html

The thin film resistive element is low tempco and low current noise, but they're magnetic, the electrode contains nickel.
RR 0805 series is cheap

Jocko, what (current production) resistors do you recommend for audio?
DALE CMF or ERC? Others?
Is there any good resistor available at Digikey?

Bye
Roberto

_________________
Ciao
Roberto

Last week, I was looking for real precision resistors at Digikey. I was prepared to
use tru-hole ones, 'cause it was for an application where it really counts:
drain resistors for the input cascodes of a very low noise differential amp
where they determine CMRR and the gain setting resistors where I wanted
1000.0x and not 1001.0 or 999.5.
This was quite disappointing. The best I could find were 0603 thin film,
though surprisingly accurate for the size.

I don't give a damn if the bias current of a LED or the trimming of a
band gap is magnetically contaminated. On the other hand I would not
accept a Cu-Ni-Au plated board for microwave striplines because it
makes a demonstratable effect.

What bothers me with Jungs reference is the large feedback capacitor
from output to the input of the op amp. I wonder if it could degrade the
amp input if the output of the REF is shorted. (20 mA limit of the protection diodes).


Gerhard

_________________
Everything has been said already - but not yet by everyone. (Karl Valentin)

I don't know what to recommend today. Resista became Draloric, and Draloric is now.......????......I dunno what Vishay is doing now. Maybe Charlie Hansen knows the answer.

I have never been fond of Dale resistors. We used Corning a long time ago, and switched to Resista when Corning vanished. They all have steel end caps, so getting away from magnetic materials is virtually impossible.

Jocko

I've been using the Vishay PTF56 and PTF65. Non magnetic, .1%, low ttc.

A little dry sounding but most everything I do has an input and/or output transformer in it which helps to counteract it.

The noise floor on my FFT is in the 2 nV/rt-Hz range. So, this is around that level. Quiet enough for government work, and probably quiet enough for DIY work.

The plot is made using some program that Gordon sent me. It clips part of the plot, so you are going to have to fill in some of the info on your own. Most of it should be self-explanatory. If not, speak up.

Yes, this uses the mystery op-amp. I'll make one with a '5534 in a while.



Jocko

Here is a '5534. Not bad for something found in the bottom of the junque boit. Not as quiet, but if it is all that you have, should work in 90% of your applications.



Oh, these were done with 2 LEDs, and no divider. Output is 3.8 V, and 7.6 mA.

Jocko

Maybe some better lighting would help. Maybe not.



Jocko

Vishay bought Draloric. They had earlier bought Roderstein, the orginal makers of Resista's.
Parts Connexion carries the Dralorics, but I have not tried them.
http://www.partsconnexion.com/catalog/resistors.html

Percy carries Resista's, but only has certain values, and has been replacing Resista's and Holco's with PRP's.

Randy

Yes, but I've heard that Vishay has now closed Draloric, and switched production to another company that they gobbled up.

Jocko

I believe your buddy JC is a big fan of the PRP audio line seeing as he was the choice ears that helped produce them.

They are available at parts connexion and should remain reasonably priced until such a time as they discover the leads have any copper content.

PS: I now maintain that FM's just suck. They wash out the detail and cut the high's, leaving them off pitch, what people call "warm". I'm aware it's a non issue for the intended application but it would still be interesting to see your graphs done with say NHG or even those UCC solide AL's if you could.

Nice results though.

Uh...........not giving away "trade secrets", but...........I am not using those UCC solid polymer stuff for low-noise applications. (They may be ok on the output, or at the chip.) I just used the FMs as that is what was on my bench.

I am going to look at some low-profile NHG caps. Remember, this is for digital circuits. Not sure how much the "sound" of the caps as audio bypass will translate to digital bypass.

I assume the handful of guys who want to buy proto PCBs will listen to it with different caps. I just measure it. (Bitter old fart who is losing interest in having to actually listen to how it sounds.)

Jocko