oranda
Active Member
Nice! Love the video. Just was looking at the schematics for the Cornet and the Trumpet. What are the advantages of the Trumpet's MM gain section over the Cornet?
Last edited:
The best phono preamp under $1000 (aka the Hagerman Bugle upgrade thread)
- Thread starter oranda
- Start date
Jim Hagerman
Member
Trumpet is a giant leap ahead of Cornet in terms of body, weight, and dynamics. It's all due to the differential nature of the design (twice as expensive), which affords constant current pull for power supplies (class A operation), common mode noise rejection, and less TBD. I do all my listening at home with a Trumpet MC.
TBD (transient blocking distortion) is something that happens in tube amps where large signals bump into grid current (positive swings) thereby causing a shift in DC bias across the preceding coupling capacitor. This can be seen in the Cornet second gain stage where large amplitude treble signals push past the self-bias (usually around -0.6V) into positive grid current territory. Not so obvious with continuous sinewave test, as the stage re-biases and keeps working. It's only apparent during short-term transients, where upper half of signal gets compressed, bottom half does not. My gain and EQ topology work to minimize this (overload in a Cornet starts at 160mV input), but it can still happen with pops and tics. The balanced stages in the Trumpet however, cancel each other out to a first order, in terms of TBD. You can hit it really hard without it flinching.
TBD (transient blocking distortion) is something that happens in tube amps where large signals bump into grid current (positive swings) thereby causing a shift in DC bias across the preceding coupling capacitor. This can be seen in the Cornet second gain stage where large amplitude treble signals push past the self-bias (usually around -0.6V) into positive grid current territory. Not so obvious with continuous sinewave test, as the stage re-biases and keeps working. It's only apparent during short-term transients, where upper half of signal gets compressed, bottom half does not. My gain and EQ topology work to minimize this (overload in a Cornet starts at 160mV input), but it can still happen with pops and tics. The balanced stages in the Trumpet however, cancel each other out to a first order, in terms of TBD. You can hit it really hard without it flinching.
knotscott
Super Member
My bugle3 has really stepped up in the past three days of testing and listening. But it doesn’t touch my Cornet2 yet. I’m not sure it can get there - the Cornet2 is just so darn smooth and elegant while still being able to kick butt when required. The upside of the Bugle3 journey was that the fussing about helped me to solve a number of issues, some of which helped the cornet2 as well.Pretty ironic that I just messaged Jim today asking about the Cornet.
I’ve been able to dramatically lower the noise/hum baseline in my system now - the bugle is technically quieter but the Cornet2 hum is very low level except at painfully high levels of volume control setting and still just sounds better. Yes it’s a tube stage, so it’s different - but to me it is better still. I’m waiting for the 2210s to come in to see how close the bugle3 can get ….
oranda
Active Member
Thank you for the info! I would imagine that 160mV would be difficult to achieve for even a reasonably high output cartridge. So then is it more the V/s of the transient causing the distortion?
Obviously it's different for opamps as it's not the same architecture, but am I correct in assuming that opamps with a GBW of tens of MHz and slew rates of several volts per microsecond would be pretty much immune to most types of transient distortion at audio frequencies and phono cartridge voltage levels? Interesting since opamps are inherently differential amplifiers, a feature you engineered into the Trumpet.
Hi All - tried to do a little "continuous improvement" today - blew up the TLE2426CLP (don't know how but I think it was a power supply issue), and my spare TLE2426CLP (that one I managed to short out on my own :-( ). Question - if the TLE2426CLP dies, can/does it take other components with it? I checked as many components as I could but have no way to know if I toasted any OpAmps. Visually, it looks like the Transistor was the only one that burned. Speaking of OpAmps, the 2210s finally came in today - sadly they have to wait until I see what mayhem I caused in my "continuous improvements". For reference, my 24v 500ma power supply DC plug was a bit sloppy on the standard jack - so I changed the standard jack and somehow must have caused an intermittent/ short which toasted the TLE2426CLP Transistor.
After the first burn/smoke event, testing showed pin4 at 24.2vdc and pin8 at 0vdc, which led me to believe the transistor was the wounded warrior. So removed the dead one, installed, tested and got pin4 at 0vdc and pin8 at 24.2vdc. Then accidentally shorted the transistor for burn/smoke event number 2. I noticed that the PCB hole on the face for the left pin (Pin1)? of TLE2426CLP looks like it doesn't have any pad anymore, but when I tested it, it goes to ground as per PCB. But wondering why the new transistor, when installed, changed the voltage check results in that way as noted above. I used a heat sink for transistor installation so I don't think I damaged it during installation. But when new TLE2426CLPs arrive, I will be more cautious.
Before I blew it up, it was sounding very, very nice
Really, when things are working or mostly working, I really should just LITFA ..... and just enjoy it.
After the first burn/smoke event, testing showed pin4 at 24.2vdc and pin8 at 0vdc, which led me to believe the transistor was the wounded warrior. So removed the dead one, installed, tested and got pin4 at 0vdc and pin8 at 24.2vdc. Then accidentally shorted the transistor for burn/smoke event number 2. I noticed that the PCB hole on the face for the left pin (Pin1)? of TLE2426CLP looks like it doesn't have any pad anymore, but when I tested it, it goes to ground as per PCB. But wondering why the new transistor, when installed, changed the voltage check results in that way as noted above. I used a heat sink for transistor installation so I don't think I damaged it during installation. But when new TLE2426CLPs arrive, I will be more cautious.
Before I blew it up, it was sounding very, very nice
Really, when things are working or mostly working, I really should just LITFA ..... and just enjoy it.
Last edited:
Jim Hagerman
Member
Usually no. The problem is that your 24Vdc power supply has negative terminal tied to ground. Must be floating!!!
Jim Hagerman
Member
No, it's not a matter of bandwidth or slew rate. Opamps will clip hard on power supply rails. Tube circuits are powered by hundreds of volts, so the output signal can just keep going (by comparison). The Cornet3 for example can pump out 24Vrms at 1kHz when overload starts to kick in. That's almost 70Vpp...
Thanks Jim - as it was fine before I changed the power jack, I must have missed or did something different during the install. Good news about the low probability of collateral damage - appreciate it.
knotscott
Super Member
That's logical advice, but damn hard for some of us to follow! LOL. Some of us who have that DIY gene, also have that "oops, I pushed it too far" gene. (OIPITF)
With a large family, and little audio budget, I can't imagine ever going out and actually purchasing the kind of sound I'm getting now. That "OIPITF" gene is responsible for some french fried parts, but is also responsible for the system I have today. It's how those of us who approach electronics like a linebacker learn. For every setback, there are several leaps forward. Sorry about your toasted parts, but you'll rebound!
Last edited:
Jim Hagerman
Member
Exactly. Topology makes a lot of difference. The passive-split EQ of the Bugle2 is pretty much worst case when it comes to overload capability. Noise floor too. And yet I chose it anyway because of the sonic "naturalness" that can be achieved.
So let's take a look at overload in the Bugle2. The signal coming off the record has very loud treble and very quiet bass. Roughly 100 to 1 ratio. Let's assume we need 20dB overload margin above 5mV (1kHz) for a good phonostage. That will allow for pops and tics and dynamic musical transients. 20dB above 5mV is 50mV, and treble is roughly 20dB above that thanks to RIAA equalization, so we have effectively 500mV coming into the first stage (above 10kHz).
First gain stage is 21dB so we end up about 5Vrms. That is pretty much all that first opamp can handle before it starts clipping on power supply rails! Now we drive the first part of EQ (which saves the day). It's a 50Hz pole into a 500Hz zero. Basically, all upper frequencies get attenuated. That 10kHz overload is now back at 500mV driving second opamp. Same thing happens and output at high frequency is again 5V. This drives second equalization filter dropping almost 20dB again, presenting 500mV into final opamp. At 20dB gain it spits out 5V nice and clean.
Remember, this is for a 50mV (ref 1kHz) input signal. At 5mV, we have 20dB of clean headroom with an equalized output with all frequencies at 500mV. That is why "gain" is specified at 40dB. This is a good signal level to feed into your linestage. Keep in mind however, the transient overload issue now becomes the problem of the linestage. Can it handle the 5V cleanly? Usually they can! Especially with volume control not at maximum.
Yup. With opamps this can make a huge difference. If that transient overload signal is faster than the opamp can handle, it slews, meaning it has gone open loop! The opamp will eventually catch up and recover with the feedback loop engaging again, but at what cost? How many non-musical artifacts did it just add to the signal?
I believe pops and tics are a very good performance test. How do they sound on your system? Do the electronics ride right through as if nothing happened? A small bump in the road? Or did it turn that scratch into something far more explosive?
I’m ok with a few toasted parts along the way - as long as it’s not some stupid “expensive” boutique op amps And of course now the bugle retains a lovely aroma of failure, hopefully masked by success soon.That's logical advice, but damn hard for some of us to follow! LOL. Some of us who have that DIY gene, also have that "oops, I pushed it too far" gene. (OIPITF)
With a large family, and little audio budget, I can't imagine ever going out and actually purchasing the kind of sound I'm getting now. That "OIPITF" gene is responsible for some french fried parts, but is also responsible for the system I have today. It's how those of us who approach electronics like a linebacker learn. For every setback, there are several leaps forward. Sorry about your toasted parts, but you'll rebound!
I once rescued a 1968 Fender Super Reverb guitar amp. A beast if there ever was one. Took me about a year to get it back to functional and still have it even though it’s too loud for any sensible practice without attenuation assistance. Apparently it was a “house amp” in a bar for many years. Once it warms up- even to this day- it has a faint aroma of smoke, some kind of whiskey and some cheap perfume (I think). But it’s a surreal expertise after so many years. And yes, I blew up a few parts on that rebuild - luckily they weren’t the unobtainium parts
Jim Hagerman
Member
The long version... 
oranda
Active Member
I know it's been the darling of some very good designers on DIY audio for a while. Getting a little off topic, its surprises me a bit when I see single opamp, active RIAA filter phono preamps in some very high dollar gear. McIntosh used that design for a long time in their solid state preamps, though I'm not sure if that's still the case. I think Graham Slee does the same. In these instances, they don't sound bad (even pretty good in some cases), but I would guess that there are a number of limitations with that design.
Ah, OK, I forget that MM carts are spec'ed as outputting up to 3-8mV @1kHz. So at 20kHz with a 20dB overload margin, yes that 1/2 volt is potentially a pretty hefty signal that's hitting the front end of the phono stage. Thanks for the reminder and excellent breakdown!
I was just looking at the input common-mode voltage range on the OPA2210 data sheet, which is 3V. That's about another 15dB of gain over 0.5mV before clipping, which gives effectively a 35dB overload margin. Is there a situation where this would not be sufficient on the input in the phono stage? Are pops and clicks covered here or is there still an advantage to the headroom offered in a tube designed phono pre like the Cornet?
Asking me? I would say they are noticeable bumps in the road. Maybe this is where a design like the Trumpet really shines?
Not sure what Rega uses in their upper end phono preamps, but they are solid state designs and to my ear they seem to handle pops and clicks very well. I guess that would indicate that they've designed their phono sections with significant headroom?
Last edited:
oranda
Active Member
So... what improvements did the Bugle "Elite" have over the Bugle 2/3?The long version...
Jim Hagerman
Member
Plenty of margin for a 500mV input overload signal. The problem isn't overloading the input section of opamp (it is with tubes), it's the output. 5Vrms is 14Vpp, something that any decent opamp with +/-10V supplies can handle. But not much more! Output will hard clip before input has issues, at least with the gain distribution found in the BUGLE.
Jim Hagerman
Member
So... what improvements did the Bugle "Elite" have over the Bugle 2/3?
A nice metal box, excellent RCA jacks, and PPFMX capacitors. Same resistors and opamps.
Well - despite my best efforts to kill the bugle3, it’s now officially done
Got the replacement parts all in place, voltages looked good. Put the original op amps in and it worked. Let it settle and play for a few hours and then took the brave step to replace the op amps with 3 2210s. Let those settle in for a few hours and then replace u4 with the opa 1612.
Such a great outcome and surely a testament to a great design that even the rookie diy warrior could not kill it - thanks to all for the support everyone and to Jim for giving us such great products to try and break .
I cannot get over how quiet this device is. I managed to significantly reduce my noise/him floor through this process. But this “final product” is even quieter than before and than expected.
Enjoy your music - I know my enjoyment level has improved greatly over the recent experience.
Got the replacement parts all in place, voltages looked good. Put the original op amps in and it worked. Let it settle and play for a few hours and then took the brave step to replace the op amps with 3 2210s. Let those settle in for a few hours and then replace u4 with the opa 1612.
Such a great outcome and surely a testament to a great design that even the rookie diy warrior could not kill it - thanks to all for the support everyone and to Jim for giving us such great products to try and break .
I cannot get over how quiet this device is. I managed to significantly reduce my noise/him floor through this process. But this “final product” is even quieter than before and than expected.
Enjoy your music - I know my enjoyment level has improved greatly over the recent experience.
desertrat748
Super Member
Jim,
When building a bugle2, would it help with clipping if it was built with lower gain in the 30 - 34 dB range rather than higher 40+ range?