I strongly recommend that you urge your DM to ditch fumble and critical tables altogether. It sounds like this will be difficult to do, but you’ll find the game much improved – and, I think, so will he. This is because...
Critical and fumble tables have myriad problems
The Dungeon Master’s Guide even strongly recommends against them, for good reasons – and there are yet more reasons to be found beyond what the DMG offers. In many ways, they are similar to the deck of many things spread out over a longer period of time. They destabilize the game and, sooner or later, are likely to contribute significantly to its end – and rarely in a positive way.
Some of the problems:
Increased swinginess – because the average PC sees far more rolls than the average NPC, they are far more likely to suffer disaster from a critical (against them) or fumble (of their own) than the average NPC. This despite the fact that the average PC is far more important to the story and game than the average NPC.
Individual rolls are more important – because warriors in 3.5 roll more dice (iterative attacks, rather than a single saving throw), they are far more likely to fumble than spellcasters, even though spellcasters were already much more powerful. Worse, being a high-level warrior makes you more likely to fumble, not less.
Context ignorant – criticals and fumbles are, obviously, random. That means they increase the chance of random death – a disappointing, anticlimactic situation that disrupts the flow of the game. This is what’s known as “boring failure.” Even when people don’t die, a fumble just stops them from doing what they were doing – still boring failure, though less dire.
Limited benefit – the results of a critical or fumble rarely add much to the game, because they don’t tie in to the plot, the characters, or the tactics that are in play. They just randomly happen. As such, they are extremely limited in their ability to produce truly interesting results.
All of these things yield a rather uninteresting and problematic situation. The default critical and fumble rules are not exactly fascinating, but they at least are fairly limited in how much “space” they take – they’re kind of boring, but they’re quick and easy boring. The small interest they provide is equitable with the small cost they require, in terms of risks, time, and effort. Critical and fumble tables are not.
Don’t believe me? I’ve written a huge amount on why these things are problematic; that material may be useful in convincing others.
A solution – disregard tables, acquire Fate Points
The solution to these problems is to eliminate critical and fumble tables altogether. They don’t accomplish their supposed goal – inject a little more opportunity for interesting things to happen – and in fact reduce opportunities for interesting things to happen. The auto-miss/auto-success rules are fairly important for mathematical reasons – they put diminishing returns on attack bonuses and AC, saving throws and DCs. That’s good enough for random roll-effects.
That doesn’t mean that the thing people usually want critical and fumble tables for is bad; it’s not. It’s just that critical and fumble tables are a bad way to accomplish it. Here’s a thought on a replacement: shamelessly steal from Fate, which is all about allowing “interesting failure.” Fate has a system that revolves around aspects that are invoked or compelled – it can be integrated into 3.5 (I’ve played such a mish-mash; weird, but it was fun), but I’m aiming for something simpler and more tied into the notion of criticals and fumbles.
Imagine if, when you roll a nat-1, the DM could offer you something in return for a fumble. Specifically, “roll on the fumble table, and I’ll give you a Fate Point,” (hey, if we’re shamelessly stealing, might as well be honest about it). Or, when someone rolls a nat-20 against you, “you can get a Fate Point if you let him roll on the critical table.” Now when you roll a nat-20, or someone rolls a nat-1 against you, you can spend a Fate Point to roll on the critical table yourself, or have them roll on the fumble table. This allows you a modicum of control over where the criticals and fumbles happen – and gives the DM reason to avoid having you critical or fumble too often, since sitting on a pile of Fate Points potentially makes you rather dangerous.
That eliminates a lot of the mathematical problems with criticals and fumbles. But, they’re still kind of boring; they’re still some random thing off a table. To improve on this, again, the solution is to ditch the tables. Instead of “roll on the table,” your DM can just suggest a fumble or critical that’s thematically appropriate. He could pick it from the table if he needs ideas, but even better if he comes up with something himself. And then you can suggest a critical or a fumble when you spend your own Fate Point. For this version, I recommend that your DM can choose to give you another Fate Point instead of accepting your Fate Point; that way you cannot suggest some ridiculous critical or fumble (or he’ll just hand you a Fate Point and tell you to just roll normally), forcing you to find a reasonable critical or fumble.
Because you are choosing these fumbles or criticals at the time they happen, you are freed to make truly interesting suggestions, that take into account the current situation, the characters involved, the plot, and so on. This eliminates the arbitrary nature of criticals and fumbles, and allows you to potentially tie these things tightly into the story itself.
To really make it like Fate, you could add one more layer: Refresh. Basically, you start every play session with some minimum number of Fate Points (or however many you had last time, if you had more). Then, in order to not fumble or be crit, you have to pay a Fate Point. So sooner or later, you have to let those happen – but then you get a Fate Point to avoid the next one. By the same token, the DM can choose not to have NPCs fumble or take a special crit, but he has to use Fate Points from a shared pool for NPCs to do it, and you get that Fate Point. At this point, this is pretty similar to how we played our 3.5+Fate mash-up game (we also used Fate for skills), and... it worked pretty well.