Hi Stefan, yes they look a bit more practical ! If I can add them to my test that would be very nice indeed
power supplies are no problem
Channel 0 of my Timepod was the worst one and we suspected a
broken input attenuator chip. I have finally repaired that after a year
and it seems it was a success.
But my references are two Morion MV-89A at 10 MHz and the
Timepod production verification test script expects 5 MHz. I felt
not confident to change the script. So I decided to measure some
oscillators that might pass as a reference.
The 5 MHz units are all from butchered Lucent KS24361 GPS
receiver redundancy units, built by HP/Agilent for Lucent.
They were spare parts still in the original box, never spent 20
years on a telecom tower in China.
The redundancy units have no GPS receiver boards, so I got
them for a relatively good price. Shipping from Florida to .de
cost an arm and a leg, however.
They were OFF since the production end test maybe 20 years ago.
If they run for a month, they might improve, esp. closer to the carrier.
Unit 105 was significantly worse than the others, but after running for
a night it caught up. Maybe it has vibrated some debris from the crystal.
The Morion MV89A I got from China had a low success rate.
Power supply was a Rohde&Schwarz NGT 20 , with some low
pass filter in the Hammond cast alu test box.
Unit 100 ( 5 MHz MTI-260) and 200 ( 10 MHZ Morion) are permanently
mounted in a Hammond box with a Mean Well 230Vac to 12Vdc
switching converter. I added 1000uF on the output but not much more.
That does not seem to be much of a disadvantage. Unit 100 and 200
have been on for a week or two, that's why they have their own switcher.
In fact, unit 100 is the best of the 5 MHz units.
The MTI-260s are mostly the same, that's large series production.
There are rubidium-like versions of the 260, but they do not go
into Telecom towers.
broken input attenuator chip. I have finally repaired that after a year
and it seems it was a success.
But my references are two Morion MV-89A at 10 MHz and the
Timepod production verification test script expects 5 MHz. I felt
not confident to change the script. So I decided to measure some
oscillators that might pass as a reference.
The 5 MHz units are all from butchered Lucent KS24361 GPS
receiver redundancy units, built by HP/Agilent for Lucent.
They were spare parts still in the original box, never spent 20
years on a telecom tower in China.
The redundancy units have no GPS receiver boards, so I got
them for a relatively good price. Shipping from Florida to .de
cost an arm and a leg, however.
They were OFF since the production end test maybe 20 years ago.
If they run for a month, they might improve, esp. closer to the carrier.
Unit 105 was significantly worse than the others, but after running for
a night it caught up. Maybe it has vibrated some debris from the crystal.
The Morion MV89A I got from China had a low success rate.
Power supply was a Rohde&Schwarz NGT 20 , with some low
pass filter in the Hammond cast alu test box.
Unit 100 ( 5 MHz MTI-260) and 200 ( 10 MHZ Morion) are permanently
mounted in a Hammond box with a Mean Well 230Vac to 12Vdc
switching converter. I added 1000uF on the output but not much more.
That does not seem to be much of a disadvantage. Unit 100 and 200
have been on for a week or two, that's why they have their own switcher.
In fact, unit 100 is the best of the 5 MHz units.
The MTI-260s are mostly the same, that's large series production.
There are rubidium-like versions of the 260, but they do not go
into Telecom towers.
Attachments
10 and 15 MHz
Piezo203 is the worst, I bought it on the flea market at the VHF meeting in Weinheim.
I got piezo200 for free as an add-on, it looks awful, but at least far out it is
the best of the bunch.
The Lucent traces are not running free but are locked to GPS.
Lucent_15 MHz is the normal output of the GPS receiver, the tripled MTI-260.
Lucent _10 MHz is one of the 4 10 MHz outputs of my add-on.
It has a CFB amplifier after the oven, then a doubler with 2 BF862,
then a CFB amplifier for each output.
< http://www.hoffmann-hochfrequenz.de/downloads/DoubDist.pdf >
Piezo203 is the worst, I bought it on the flea market at the VHF meeting in Weinheim.
I got piezo200 for free as an add-on, it looks awful, but at least far out it is
the best of the bunch.
The Lucent traces are not running free but are locked to GPS.
Lucent_15 MHz is the normal output of the GPS receiver, the tripled MTI-260.
Lucent _10 MHz is one of the 4 10 MHz outputs of my add-on.
It has a CFB amplifier after the oven, then a doubler with 2 BF862,
then a CFB amplifier for each output.
< http://www.hoffmann-hochfrequenz.de/downloads/DoubDist.pdf >
Attachments
Last edited:
Hello everybody! So, it's time to write my feedback. I've been using Andrea's clock for a couple of months now. Before that, my favorite was the tube-clock from Abbas. Tube clock allowed me for the first time to understand that digital sound can be indistinguishable from the analog. Solutions like NDK or Crystek are really, just toys, it was not possible to achieve the proper sound level. Oddly enough, but the most difficult for the playback remains the CDDA format. It comes to the point of absurdity when modern DACs can only play hi-res with sufficient quality. But I know that CDDA is not an easy nut to crack and I need to keep digging. Well, okay, I'm distracted! ..
I have a version of Drixo 11.2896MHz, also powered by Andrea's TWRPS-UGL-F. The signal is fed to the FSDO-S converter, and then to the fifopi, and then to my beloved TDA1541A. The Clock took a while to warm up. But now, when it never turns off in my system, I am completely satisfied with the result! Now I can safely compare digital sound with very cool vinyl systems, no doubts!
The sound from my DAC from Andrea's clock seemed so good to me that I recorded some audiophile recordings on to the Benchmark ADC1. During the repeated playback, it is very difficult to distinguish the recording from the original, on the verge, I would say.
Let me say: BRAVO Andrea! The Very Well Tempered Job!
I have a version of Drixo 11.2896MHz, also powered by Andrea's TWRPS-UGL-F. The signal is fed to the FSDO-S converter, and then to the fifopi, and then to my beloved TDA1541A. The Clock took a while to warm up. But now, when it never turns off in my system, I am completely satisfied with the result! Now I can safely compare digital sound with very cool vinyl systems, no doubts!
The sound from my DAC from Andrea's clock seemed so good to me that I recorded some audiophile recordings on to the Benchmark ADC1. During the repeated playback, it is very difficult to distinguish the recording from the original, on the verge, I would say.
Let me say: BRAVO Andrea! The Very Well Tempered Job!
Within the few last years I have heard the one only pure analog (vinyl) system, the sound of which has really impressed me. Mainly due to unbelievably realistically scaring dynamic range. But the cost and complexity of the whole system multiplied by the cost and constant maintenance needs, just informed me: you not ready for such, forget it for a while. Do I want the vinyl system? Of course! Do I have a possibility right now for what is needed for that? Not now at least.
BTW, if somebody have a wish, I can re-record any CDDA track, for your own judgement in your system. But please take in mind that it will sound like the original track at a maximum, I can't make it better than original. And due to 11MHz, I am limited by 24/44 format.
BTW, if somebody have a wish, I can re-record any CDDA track, for your own judgement in your system. But please take in mind that it will sound like the original track at a maximum, I can't make it better than original. And due to 11MHz, I am limited by 24/44 format.
The best SQ phono is from direct-to-disk-lathe recordings. No tape, no digital. The reason my friend repeatedly insisted I need to have some kind analog playback capability is for comparison with the sound I work on trying to get from dacs. The phono is a kind of a repeatable reference that is not digital. Also, have some real instruments here. Real can be compared to phono compared to digital. The point is whether or not the dac sound is moving in a direction that sounds closer to real, or is it moving away from that in way that fools the brain into merely thinking its getting more real. Measurements help too, of course. IMHO it all helps for judging progress.
Last edited:
Sorry Andrea, this has the potential of going extremely off topic - vinyl versus digital
Allow me one comment and I will stop, promise….
True, vinyl can still beat digital in the mentioned aspects, BUT you have to spend crazy money. My Garard, linn lp12, tonearms, amplifiers and cartridges costed me more than 30k. And that is far from what you can actually spend. Enough room for more
When spendings for digital and analog are similar I would agree both play on par….
Sounds like a good approach particularly if you are building a product. I kind of get the impression you are in this professionally?
I certainly appreciate when my digital/triode based illusion convinces me it is more natural. Sometimes its also good as a consumer to accept the illusion and not get confused by the reality of just how unnatural it actually is. At the end of the day I don't have a chamber orchestra in my room, I just have my system.
Some of my friends build products. I am only a serious hobbyist so far. If that ever changes I will say so here in the forum so people can take it into account as they see fit.
Even if I ever did design a dac product, I told my friends I would never, ever make back more than I put into it. They laughed; they know its true. I wouldn't do it unless it sounded really good. In that case it would be expensive to make and thus would have a very low sales volume.
Both of which predate CDs.
I have some 1920s direct to disk recordings that are wonderful, but certainly not the best SQ. can anyone name me anything from Sheffield labs or Crystal clear that warrants a second listen?
Rubbish music is only of interest to the stereotypical audiophile joke. The rest of us are music lovers.
- Status
- Not open for further replies.