Zen -> Cen -> Sen, evolution of a minimalistic IV Converter

[qusp]

ive sent off a request for a quote on the k369-V at utsource, but i would still like to finnd some prematched

[EUVL]

I don't think you can easily find anyone selling 2SK369V quads or octets.
If you buy enough and match, I am sure you have NO problems getting rid of them with the GB gang.


Patrick

[qusp]

yeah the only source of matched i found was an OLD thread on here that hasnt been visited since mid 2010, i sent the OP a message as in his last posts he did seem to be indicating quite a few left, but at 22eu per quad back then (and he did seem to have 4 closely bunched quads) not exactly a cheap option. so probably what i'll do is first test CEN with some 11ma J74/k170 and make up the SEN with k170 initially for setup and if all goes well; move onto the k369.

I just dont have the time to dedicate to matching this time of year; especially when i have most of the devices needed with the k170 already matched and other unmatched parts to find the rest in, its hard enough finding time for DIY as it is

you actually posted in his thread asking if there was a P channel to match, it was that long ago ha.

i'm sure i'll do it, it'll keep eating at me til i do, but its going to have to wait if i'm going to get this initial testing done for the 9018/9012 gang anytime soon

i'm going to have to make another order for some more heatsinks from Mark for more testing, are you getting any of the quad sinks in again soon, or should i just go ahead and order more of the 2 device parts?

[bcarso]

On the subject of matching, a few cents worth: you can parallel JFETs with less-than-perfect matching and get a composite that way that more-or-less averages things out. So if you have a big bag of, say, SK170 parts, ideally with at least the same date code, you may find that a simple presort based upon on-resistance, and then paralleling a bunch for each "element" of the given circuit suffices. The gotchas include the propensity for parasitic RF oscillations with long lead lengths and traces. One option to combat these I've been looking at is, depending of course on the specific circuit, a strategically placed resistor in each drain lead can serve double duty of reducing signal-induced self-heating, the so-called "thermals", while also damping the circuit interconnection parasitics. Feucht has some good material on thermal distortion in his book Handbook of Analog Circuit Design; the past masters who addressed these concerns were oscilloscope vertical amplifier designers, although they didn't publish much

Since I have a goodly number of 2SK364BL parts from the detritus of a Harman project shutdown many years back, I may put this recommendation to a test. The 364, by the way, is the same chip as the SK170, just characterized for analog switching.


Brad

[ryanj]

Hi Boys,

I've been through a few 9V block batteries with the Cen now, and with my setup I've noticed that i need a considerable amount of headroom to keep the high amplitude signals from clipping. I'm using a 1.5k IV resistor with +-2mA current swing (TDA1541). Once the voltage went below 17V it sounded crap. To say the least. Bass started sounding weak and less dynamics overall. So now i have about 20 x 9V batteries with 8.4V.

After my wedding was over on Saturday i finally had time to put some rechargeable lithium batteries in. I've now got a bit over 19V available now, and after a short listen it sounded very good.

Before i got busy with organizing and preparing for the big day i had time to get a different current buffer working to supply the required +2mA using a Toshiba k216 mosfet. I found that there was much less thermal drifting, but still, depending on ambient temp i could have between +-15mV drift. After i found out from John Brown that he measured better performance with the 1541 at about -45mV i settled to let it drift around this figure. But because of DC coupling to my F5 i had to re-bias and get the DC on the output close to 0 at my favorite listening level.

Hope to hear some listening impressions of the Cen from other people that are using the TDA1541 soon!


bat.jpg dacbat.jpg
Yes, it looks messy, I'm making a proper board soon.

[EUVL]

> I've now got a bit over 19V available now, and after a short listen it sounded very good.

Good to know.


Patrick

[mfrimu]

Hello,

The heatsinks for the SEN are not the same layout as the PCB. At first glance, the PCB looks correct.

Heatsink image attached.

Thanks.

[EUVL]

Mark sent you the wrong ones !!

You should have gotten Type 10 :

xen-audio

If many or all of you have the same, please report at the GB thread.
We'll send you new ones at our costs. Takes a few weeks to get them made.

SEN / CEN All JFET IV Converter Evaluation Pack

If you wouldn't mind, we should be very grateful if you could put these 4x type 11 sinks in an envelope and send it back to Mark by ordinary mail. Unless you want to buy them off for using in a 4-JFET CEN for half the Zin and double the bias. We'll offer you a 20% discount in that case.

Terribly sorry for the mistake. we are all getting old and useless ......


Patrick

[samoloko]

I got type 11 heatsinks too
but as I see pcb layout type 11 Is relative to CEN and type 10 Is relative to SEN
or I am wrong

[qusp]

thats correct. as i said in the other thread, i'll buy your type 11 as i have the 9018 and got 2 each of the 2 device to make up each type 11, thats fine and will work no worries, but type 11 would be ideal. i have already contacted mark about buying some extras as i need some for other projects, but i dont really need as many as the minimum order, so just buying yours would be cool. i'll follow through with Mark if need be, but if this presents a nice solution i'll buy yours and pay shipping to me. seems like win win