Showing 1 of 1303 conversations about:
View Full Discussion This update is for members who joined the drop in March:
The Massdrop x THX AAA™ 789 Linear Amplifier is making its way through production! We also have some important updates to share about the gain specifications.
We worked with THX to adjust the gains for the primary use case of the Audiophile community. Previously, the gains were optimized for variance in source output power. Most Audiophiles typically use 2V DAC sources and wish to drive in-ear monitors or hard-to-drive headphones.
With these slight changes there were some updates to the gain specifications, which we have listed below:
- Frequency response: + 0.01 dB / - 0.03 dB 20 Hz–20 kHz; + 0.05 dB / - 0.15 dB 10 Hz–50 kHz
- THD: -130 dB, 32 ohms, 1 mW: 0.000032%
- THD: -130 dB, 32 ohms, 100 mW: 0.000032%
- THD: -122 dB, 16 ohms, 1 mW: 0.000080%
- THD: -122 dB, 16 ohms, 100 mW: 0.000080%
- Gain: 0.66x , 2.0x, 6.6x (-4, +6, +16 dB), selectable via front switch
We’re on track to meet our estimated ship date of October 31, and we'll check back in around the middle of October with the status of the shipment.
- Output power: 1800 mW, 32 ohms, <1% THD (watts per channel)
- THD: -135 dB, 300 ohms, 100 mW: 0.000018%
- THD: -133dB, 32 ohms, 100 mW: 0.000022%
- THD: -130 dB, 16 ohms, 1 mW: 0.000032%
- THD: -125 dB, 16 ohms, 100 mW: 0.000056%
- Gain: 0.33x , 1.0x, 3.3x (-10, 0, +10 dB), selectable via front switch
> Gain: 0.662x , 2.08x, 6.616x (-4+6, +618, +1624 dB), selectable via front switch
I'm sorry, am I reading right that the Balanced output allows one-thousand six-hundred and twenty-four decibels of gain? Are we trying to distort spacetime with our headphones?
" Are we trying to distort spacetime with our headphones?"
Yes, we are... But I do think it's a copy+paste error where the decimal points were dropped. I don't even think the tech exists that can measure -13 kdB, and this amp certainly can't output 18 kW
@Shynd good catch! There were some errors in there - apologies! The specs have been updated.
Error in mine, too. No such thing as a kilodecibel. I should have written "hB"
Great news, thanks for the update.
The new gain settings are very much welcome, and the lower THD for low loads is great news!
Hi Avery, so for this drop as well as future ones, will the update apply as well?
As the original drop didn't hit 480, will the buyers up to 480 be guaranteed a sub 500 serial number as with the original drop? I also understand that there are some units in this batch that are already in production but didn't sell from the first drop (as CEE_TEE stated on head-fi), will the early buyers be able to get the original drop ship date?
So does this apply to current drop too?
Any updates to the max dBu in level before clipping?
For those of us not using difficult to drive headphones, this shouldn't cause any issues correct?
I think the adjustment (lowered gain) is to accommodate easy-to-drive/sensitive earphones and headphones, so no issues. If anything, you should rejoice! :)
Thanks for the update :)
What about those of us that have sources with 4V or 5V outputs? How does this change affect us?
Also, what are the output power specs now?
Client of the first drop here.
The 789 can tolerate the following rms input voltage from your DAC without clipping:
-10dB Mode: 7 V
0 dB Mode: 7 V
+10 dB Mode: 2.1 V
And on the 789, the RCA unbal and XLR-3 bal have same input voltage tolerance.
So for my DAC that outputs 4.5Vrms, which of the following is correct?
- I should make sure not to use the highest gain
- I can use the highest gain without going past a certain volume
Here are two good options for you that will avoid clipping at any volume level with your 4.5V DAC:
1) Use 789 in 0dB or -10dB mode all the time and you'll never clip even at full volume. You'll achieve 4.5V for your TRS headphones, and 9V for your XLR-4 balanced headphones.
2) To access the maximum voltage swing... attenuate 7dB in your player (e.g. Foobar) or DAC, and then use the 789 in +10dB mode. That'll achieve 7V at your TRS or 14V at your XLR-4, and never clip at any volume level.
Thanks for the quick and detailed replies.
Just to confirm, are you implying that I should avoid the maximum gain altogether without any form of prior attenuation? Or I can still use it by making sure to select a lower volume on the amplifier's knob?
Hi citraian, yes you should avoid the max gain position unless your DAC is guaranteed (e.g. with attenuation) to not be putting out more than 2.1V. Else you'll clip the preamp inside the 789.
I am confused now, does that means the maximum output power cannot be reached with the lower gain setting?
Correct. The -10dB setting is mainly for sensitive IEM/headphones, where lowest noise and distortion and vol knob usable range are the goals, not 1000's of mW of output power. Here, the max output voltage/power are also reduced by 10dB.
If you can't get adequate SPL at -10dB, then switch the gain knob to the 0dB (or even the +10dB setting if you have <2.1V DAC). You can safely change the gain knob -10/0/+10 realtime without any clicks/pops.
I must say I am a bit disappointed, as I thought the XLR input could handle pro levels without issue (typically up to +24dBu at 0dBFS).
If 7V/19dBu is the clipping limit (1% distortion?), then what max input voltage should be considered to keep the stellar performance of the amp unaffected?
The 7V/+19dBu output is actually unclipped. Clipping (1% THD) starts a bit higher than that.
I don't think we ever released any specs stating support for +24 dBu sources. Supporting +24dBu input would have required a 4th position on the gain knob and more relays internally. Due to CBOM budget limits, something had to give.
Thanks for the answer, I understand. It is just that the new gain specs had my hopes high.
I guess I will have to use a passive fixed attenuator between my DAC and amp.
How is the input impedance on the XLR? Will it behave correctly (CMR, frequency response, etc.) with a passive pad in front of it? (I have had bad experiences with anaview amps in the past).
XLR Input impedance is 50k differentially (line-line) and 1Meg each line to GND.
You could build a 2:1 R divider for each line to gnd to attenuate 6dB. E.g stereo balanced with total eight 600 ohm 0.1% 1/4W metal film or 1206 thin film resistors.
Or preferably attenuate 5dB digitally in your player e.g. Foobar /JRiver / etc.
It's a strange design decision since this doesn't pair well even with Massdrop's own upcoming DAC which outputs 2.5V...
I thought it would at 0 or -10dB gain setting? o_O
While I'm aware of the trade-offs involved seems an oddly low input "tolerance". With the exception of portable or low end DACs, decent ones out there are pushing 4-plus V to the amp.
This amp has no problem with 4VRMS input, it is with >7VRMS input that you clip the input stage.
So the problem only concerns DACs that have an output level higher than 19dBu at 0dBFS, were you have to resort to either digital volume control or insert a passive pad.
With a < 7VRMS/19dBu DAC you simply have to choose the appropriate gain setting.
Thanks. I misunderstood/missed some of the above. Makes much more sense now.
What you said is valid only for the low and medium gain settins. The highest gain setting clips at anything above 2.1V
2.1VRMS is actually the amp sensitivity when used with the highest gain setting, as you get very close to the 1% THD power specification.
Just curious. I'm going to be using a balanced t50rp mod (mkii argons) with this amp and seeing as they are a hard to drive headphone I would like to get as much power from this amp as possible without distortion. Would I get more power running 7V into the xlr input and only using up to medium gain, or will I get more power using 2V into the xlr input and using up to max gain? Also, is one option going to give me a cleaner signal (lower distortion and SNR) than the other (my source is a matrix quattro II)? Thanks.
Hi db42! Sounds like a nice setup. Either arrangement you mention will yield identical full output power. And do you have a 4-wire balanced to XLR-4 for your T50RP? That'd let you use the XLR-4 on this amp and give you 6dB more output power than from the TRS.
Hi Andrew. Thanks for the reply. I haven't gotten a cable yet, but I do intend to run these balanced via xlr once they come in. Can't wait to receive the amp and give it a listen.
Could you please expand on where you are in the production process? We were assured months ago that the design for this product was "locked in", and your last update presented design changes. Are additional design changes still being made? Which parts of the amp have been physically manufactured and which are still awaiting production? Another electronics drop with a 6-month lead time that I'm currently waiting for announced delays from design changes to its pcb and chassis right when it was supposed to be shipping, so I'm concerned about the process Massdrop follows. Some more insight into where you are in the process (and what exactly "on track" means) would be appreciated.
What if the DAC is 2.1 - 2.2V RMS SE, but higher (approx 2x) for balanced? Can we not use the high gain for balanced in that case?
Right. High gain can only tolerate max 2.1 Vrms input in either BAL or SE.
Functionally speaking, since I don't know of many balanced DACs that are under 2.1Vrms, For SE, you'll have potentially 3 gain settings, and in most balanced cases, you'll have two gain settings, but that should be more than offset by the high Vrms of a balanced signal to begin with.
Hi Andrew. With your measurements/observations, and given how you've "tuned" the gain variables, which combination of parameters (input voltage, gain setting, and headphone load) should yield the best performance (ie, dynamic range and lowest distortion). Thanks, Sean
Use XLR connection to the DAC if you can.
And to squeeze out every last dB of THD+N and to maximize volume pot usable range, choose the lowest 789 gain setting that gives you adequate SPL while you're listening.
HP connection: 1/4" TRS is great for low output impedance and convenience, and XLR-4 is great for high voltage swing and minimal crosstalk due to separate ground pins. Either way ensure your HP cable has separate ground wires for L and R headphone, all the way to the plug.
Don't sweat it too much though - we designed the 789 so no matter how you use it, it'll sound great.
Hey man, thanks for the quick response. Why quibble over a few .00000001's, right? :) Given all these measurements accurate, then my source is going to be making the more noise than this preamp/amp anyway.